Acute disease within the abdomen is common and many patients with abdominal symptoms present every day to doctors working in the community. Within a Western population of half a million people, between five and ten patients are admitted to a surgical ward each day with acute abdominal pain. One or two more will complain of acute abdominal symptoms after an accident. By definition the illness starts suddenly and most patients present to a hospital within 7 or 10 days of the onset of symptoms. In the majority of patients, symptoms arise from disease within the abdominal cavity itself, but occasionally they originate elsewhere in the body.

The range of disease, extends from the relatively trivial to the immediately life-threatening and attempts to reach a diagnosis must sometimes be curtailed in the interests of immediate treatment. More commonly there is time to take a history, to examine the patient, and to organize the investigations which will be helpful in establishing a diagnosis and planning treatment. Accurate recording of the relevant facts is vital and a clear understanding of the anatomy and pathophysiology of intra-abdominal disease is necessary for both diagnosis and treatment. These patients are therefore ideal for training junior members of a surgical team.

Some patients require early surgery. This itself varies from a simple straightforward procedure to a highly complex operation which stretches the ability and the skill of even the most experienced surgeon. The immediate feedback that an emergency operation provides on the accuracy and the adequacy of the preoperative assessment and preparation is another reason why the patient with an acute abdomen is an important part of surgical training.

A good knowledge of normal and abnormal abdominal anatomy, and particularly surface anatomy, is essential. Variations within and between individuals are obvious, but normal anatomy also changes with age, posture, respiration, disease, and previous surgery. Nevertheless with experience most surgeons carry a remarkably accurate mental picture of the expected internal position of any particular organ in any particular patient.

The embryological development of the abdomen is relevant in two respects. The intestine and all its associated organs such as the liver and the pancreas develop initially as midline structures. Thus visceral pain is usually felt along the midline of the abdomen. The gut also has a segmental origin so that the division into foregut, midgut, and hindgut exactly correlates with the vascular supply and correspondingly pain is felt in the epigastrium, the umbilical area, and in the hypogastrium. Certain congenital abnormalities can predispose to acute abdominal complications.

In contrast to the visceral peritoneum, the parietal peritoneum is innervated by somatic nerves. Pain is therefore accurately localized to the site of irritation of the abdominal wall and is accompanied by a reflex contraction of the abdominal wall muscles. This applies both to the anterior and to the posterior abdominal wall. Psoas spasm from acute appendicitis and a scoliosis concave to the side of intra-abdominal inflammation are two good examples. Inflammation confined to the pelvis may not, however, be accompanied by anterior abdominal muscle spasm and this may cause clinical confusion. This is because the somatic nerves which supply the organs in the pelvis do not supply the anterior abdominal wall muscles.

When describing the findings of abdominal examination the surface is best divided into six areas by a transverse line going through the umbilicus and longitudinal lines running through the tip of the ninth rib on each side (Fig. 1) 1420. Thus there are epigastric and hypogastric areas in the middle and an iliac fossa and hypochondrium laterally. It is often also useful to describe a periumbilical area. However, it is important to realise that none of these divisions have a true anatomical basis.

Normal physiology is rapidly disrupted by the onset of acute intra-abdominal disease. Many patients vomit, and gastrointestinal secretion, absorption, and motility all change in the presence of obstruction, luminal infection, or peritonitis. Urine is reduced in volume and altered in content, usually secondary to redistribution of fluid in the body compartments but sometimes because of a direct toxic effect on the kidneys.

The mediation of abdominal pain is not well understood. It is perfectly possible to handle the intra-abdominal organs and even divide the bowel of a conscious patient without causing any pain. However, distension or stretching of the bowel wall is accompanied by reflex contraction of the smooth muscle in the wall, which is immediately painful. This may be due to transient ischaemia of the muscle. The pain fibres run with the splanchnic sympathetic nerves to the spinal column, where they are distributed segmentally. The pain is localized to the abdominal cavity but not to the precise segment of bowel which is being stretched. Other pathways within the spinal column are also stimulated and vomiting, which is a common accompaniment of severe pain, can also be centrally mediated.

The gastrointestinal tract is a significant source of a wide variety of hormones. These change in response to acute disturbances of function but whether this is a primary or a secondary effect is not yet clear. Inflammation is the most common cause of acute pathology within the abdomen, followed by obstruction, haemorrhage, trauma, and ischaemia.

Bacteria, viruses, fungi, parasites, and chemicals can all cause inflammation: bacteria from the bowel, such as Escherichia coli, Streptococcus faecalis, and various anaerobes are by far the most important. Other bacteria which cause acute abdominal pain are Salmonella and Shigella species, Yersinia, and Campylobacter. Acute inflammation normally develops into clinical significance over hours rather than minutes or days, and progression either to suppuration or resolution also takes time. Perforation and ischaemia develop in minutes and cause very acute symptoms. Resolution whatever the underlying pathology always takes longer than development. Neoplasia, neurogenic, and metabolic disorders occur less commonly but they are all well-recognized causes of acute abdominal pain.

Some of these pathological processes are closely interlinked. There are a number of causes of intestinal obstruction of which neoplasia is one. Peritonitis from perforation of the bowel into the potential peritoneal space usually arises from local ischaemia, but this may in turn be caused by inflammation or obstruction which has progressed to strangulation. The clinical presentation and the physiological consequences of obstruction or peritonitis may be similar whatever the cause, but a careful history and examination should enable the underlying diagnosis to be discerned.

Most patients with an acute abdomen can be managed using simple clinical skills (Table 1) 399. An accurate history and a thorough examination are often sufficient to make a diagnosis and recommend treatment; modern investigations can help and may reassure the anaesthetist that the patient is fit for an operation. The primary objective, when the patient and the doctor first meet, is, therefore, to elicit the symptoms and the signs necessary to make a rapid and accurate diagnosis. It is sometimes obvious that the patient is in severe pain or seriously ill. The necessary immediate treatment must then take precedence over making a diagnosis.

Unfortunately, even the most experienced clinicians only make a correct clinical diagnosis of acute abdominal pain on four occasions out of five; younger doctors and those who practise in the community are only right half the time. Many attempts have been made to improve on these results, and one method which has attracted much attention is computer-assisted diagnosis. By a curious coincidence this has simply taught us once again that taking an accurate history and examining the patient carefully are still the most important factors in making a correct diagnosis.

Many patients will make their own diagnosis as one listens to their story: the art of taking a history is to induce every patient to do so. Doctor and patient have not usually met before, and the style and the approach of the doctor really does matter. A relaxed confident manner and a smile always help and you must make it absolutely plain to the patient that they have your complete attention and that you have plenty of time to listen, even if this is not so. You should discourage interruptions by other members of staff or requests to answer the telephone.

Patients like to be treated as individuals. Go and sit by their bed knowing their name, and introduce yourself clearly with your own name. Some patients immediately start to describe their symptoms and must be left to continue. Others look for a cue from the doctor. Simple non-specific questions such as ‘what has happened?’ or ‘why have you come to hospital?’ are best. Some will then give their history spontaneously; others reply in only a few words and need prompting again. It is occasionally better initially to engage the patient in conversation about something entirely unrelated, such as their job or their family, and then when they are relaxed lead the discussion back to the acute problem. This is particularly useful with very anxious patients. The most difficult patient is the one who is garrulous about everything but the reason they have come for help. Often there is nothing for it but to stop the flow of words deliberately and redirect the patient to the current problem. It is difficult to do this without appearing rude or disinterested: beware of the temptation to assume that there is little wrong with these patients. They are sometimes simply frightened.

Most patients come to the end of their story spontaneously, and sometimes they have told you everything you need to know in perfect order. Never intervene to clarify a point of detail but do stop the patient when the information they offer becomes irrelevant: it is important not to overload the brain with too many facts. When the patient has finished there will usually be some points which need amplifying or some further information which is essential. This is best obtained by asking direct but not leading questions. It is very easy indeed to suggest the answer you want either by the words you use, your facial expression, or the manner in which you speak or behave. If you do this the answers will be unreliable. Asking questions is also an art which requires tact and skill. Short specific questions are best, and they must be phrased clearly without using jargon and in language the patient understands. Some patients, like most politicians, do not answer the question they are asked. You should insist, politely, on a specific answer if one is possible. No two doctors ever obtain exactly identical histories: a younger surgeon may be amazed to hear a patient give a totally contradictory reply to an apparently identical question from a senior colleague. It is also surprising how often the very last thing the patient says clinches the diagnosis.

Not everyone can give a history themselves. Most children are shy or frightened, although others, even the very young, sometimes tell a perfect story. The confused and the mentally handicapped are often unreliable as regards facts, while the memory of an elderly patient who is ill is often faulty. A relative or a friend must then relate the history, but the clinician should remember that his or her personality then intervenes. This is a particular problem if the patient is foreign and the history has to be taken through an interpreter.

Complete attention to the patient and absolute concentration on everything he or she says and how it is said it is essential. Observation of the patient is slightly different from inspection during the examination. It encompasses demeanour as well as an assessment of personality, mood, and reaction to the illness. Movement, particularly expressive movement of the hands, is always useful. Patients with peritonitis lie quite still and look ill, patients with colic really do roll around, and patients with cholelithiasis often describe the pain radiating round into the flanks with their hands for example. Obvious and significant physical signs such as gross abdominal distension with audible borborygmi, jaundice, or the smell of melaena should not be ignored: they all point to a specific pathology which may be confirmed by specific questions.

Allowing the patient to talk freely does not prevent recording the facts in a systematic fashion. In most hospitals this has to be done freehand but there are advantages in specially designed forms. The information is recorded systematically and omissions are obvious and can be corrected at once. Such forms also require the clinician to be specific about the features of certain symptoms.

Most patients admitted with an acute abdominal problem complain of abdominal pain. Cope in his classic book observed that acute pain lasting for more than 6 h in a previously fit patient usually has a surgical cause. It is also a most important symptom: detailed enquiry about the nature of the pain will often indicate the correct diagnosis.

The first thing to establish is the precise site of the pain that the patient has now. Some patients are extraordinarily obtuse about this partly because they have difficulty in answering and partly because they often do not understand why you want to know. It is best to ask the patient to point with one finger to where the pain is worst and to record this site in the notes. Those who wave a hand vaguely everywhere probably do not have too much wrong with them. Pain often moves during the course of an illness and it is then worthwhile asking where the pain was situated at the beginning.

Radiation of the pain to other parts of the body is often diagnostic. Radiation of the pain to the testicle in ureteric colic, to the shoulders in acute cholecystitis, and to the knee with an obstructed obturator hernia are specific and typical examples. Sometimes patients volunteer that a pain radiates elsewhere but more commonly it is necessary to ask directly.

Some patients can say exactly when the pain started. They may be able to give a time or say what they were doing. This always suggests a significant cause and an acute pathological process, such as perforation or strangulation. Pain which wakes the patient up at night is also significant, although it is not often possible to describe the acuteness of onset. Sometimes pain is not the first symptom the patient noticed and this may suggest a medical cause, as with the vomiting from gastroenteritis or the marked anorexia of hepatitis. The duration of the illness gives some idea how far any pathology may have progressed and this can be correlated with findings on clinical examination.

Some patients relate the onset of their pain to an injury. Apparently mild trauma is occasionally followed by serious intra-abdominal injury; on the other hand it is more common for patients, after the onset of the symptoms, to try and relate them to an injury. This can be dangerously misleading, with acute testicular torsion for example.

There are two aspects to frequency. Alterations in the pain since this episode started are useful pointers to the immediate diagnosis, whereas pains which have come and gone in a similar way in the previous weeks or months suggest a longer-term and more chronic disease process. Variations in intensity in the short term can be classified into two types. Either the pain is constant or it comes and goes. If it comes and goes with some degree of regularity it is colic. Constant pain is associated with inflammatory conditions and colicky pain with distension of smooth muscle.

Any movement makes the pain of peritonitis worse, while lying still makes it better. Acute exacerbation of the pain on walking, breathing, coughing, or going over a bump in the road on the ride to the hospital is equivalent to rebound tenderness on examination. Pain in the shoulder on lying down comes from diaphragmatic stimulation by an irritant fluid. The fluid is often blood from an intra-abdominal injury or an ectopic pregnancy. Analgesics usually make the pain better; this can be deceiving. Sometimes vomiting temporarily relieves the pain of obstruction.

Pain is a very subjective symptom and people's reaction to it varies widely. Accompanying signs such as sweating and tachycardia give the observer some idea of severity, but this only establishes that there is something wrong with the patient which is often perfectly obvious anyway. Most patients find it very difficult to describe the nature of their pain and require prompting. No particular diagnoses are suggested by such descriptions as boring, dragging, sharp, or dull and they are best avoided.

These are two quite separate symptoms and both are useful in diagnosis. Nausea may precede vomiting but it need not do so and neither does vomiting always follow nausea. Nausea by itself is a less specific symptom, although it is a common accompaniment of gallstone disease. Anorexia is a separate and somewhat non-specific symptom since most people and particularly children lose their appetite when they are unwell. Pain normally precedes vomiting in surgical disease of the abdomen whereas the reverse is often the case in medical conditions.

Vomiting is a classic symptom of intestinal obstruction and it usually accompanies colic. Vomiting often occurs after a bout of pain in obstruction and the shorter the interval between the two the higher the obstruction. The vomit itself is initially green in colour but turns yellow and then frankly faecal as the obstruction persists. Retching without vomiting suggests acute torsion of an intra-abdominal structure.

Vomiting does not often accompany perforation of a peptic ulcer nor intra-abdominal haemorrhage, and it is a late event in distal large bowel obstruction if it occurs at all. Nausea and anorexia are more common than vomiting in appendicitis.

Diarrhoea and constipation are two potentially confusing symptoms because they mean different things to different people. It is important first to establish the patient's normal bowel habit and the normal consistency of the stool and then to decide if there have been any recent changes. Diarrhoea to some people simply means frequent defecation of normal faecal material whereas repeated loose watery stools are of greater interest to the surgeon. When true diarrhoea is present it is important to establish whether other members of the household are afflicted. The presence of blood, slime, or the black tarry stools of melaena are all of obvious diagnostic value. If intestinal obstruction is suspected then failure to pass wind as well as stool is important.

Symptoms arising from the uterus, fallopian tubes, and ovaries are a common reason for admission to hospital with acute abdominal pain. Furthermore the negative laparotomy rate is highest in young women. Questions about normal and abnormal menstrual function, vaginal discharge, and the risk of pregnancy are therefore essential. Tact and sensitivity are required but the answers really do matter: a ruptured ectopic pregnancy is a potentially lethal condition.

Alterations in the pattern of micturition suggest urinary tract disease. Frequency is linked with inflammation, while anuria is most commonly caused by acute retention in elderly men. Pain on passing urine must be separated into two classes. Abdominal pain exacerbated by micturition suggests irritation of the peritoneal surface of the bladder, while stinging pain in the urethra on urination is characteristic of infection. Patients should also be asked about the colour of the urine and the presence of blood or pus. Dysuria is a symptom which means different things to different doctors, and the term should not be used without specifying what is meant.

Any previous medical problem may be relevant to the cause of an acute admission for abdominal pain and it will certainly be relevant to the management. Chronic indigestion can be a useful pointer to a possible cause of peritonitis. A past history of abdominal surgery is important because adhesions have now overtaken hernias as the commonest cause of intestinal obstruction. Patients often report previous episodes of abdominal pain and it is useful to establish whether this episode is identical. If it is chronic surgical diseases that flare intermittently must be considered. Recurrent acute pancreatitis would be a good example.

Many people take therapeutic drugs. Most patients when asked think only of those prescribed by the doctor but in many countries in the world, including the United Kingdom, it is possible to buy drugs without a doctor's prescription and these may be relevant too. Diuretics and sympathomimetic drugs may be implicated in the onset of acute retention, digoxin overdose classically causes vomiting followed by abdominal pain, and many drugs cause cholestatic jaundice.

Not all patients know what drugs they are taking and pills may be transferred from bottle to bottle so that the labels are unreliable. Ultimately a direct enquiry to the doctor or the pharmacist who wrote or supplied the prescription may be necessary.

No experienced doctor completely separates examination from taking the history. Observation begins the moment the doctor meets the patient and does not end until they part company. Most clinicians rapidly assimilate, almost unconsciously, many features of a new patient and not all of them can easily be described in words. Attitude, alertness, mood, agitation, sweating, respiration, movement, the eyes, the colour, the facial expression, the pulse, the handshake, and many other factors are all put together to give an instant impression of the nature and severity of the illness the diagnosis. The restlessness of a patient with colic is in marked contrast to the immobility of peritonitis. The gaunt patient with sunken eyes, a weak thready pulse, and little respiratory or abdominal movement looks the same today as did patients two and a half thousand years ago when Hippocrates first described the facies of severe peritonitis.

First impressions may, of course, be false and they are not a substitute for a systematic examination. Some would say that examination does not add much to a well-taken history but more evidence to help unravel a diagnosis is usually welcome. As with the history, examination of the whole patient is relevant in the overall management, although here we are concerned with the signs which are important in the diagnosis of the acute abdomen.

Pulse rate, respiratory rate, temperature, and blood pressure are all essential observations. The initial values on admission may be misleading because of the hustle and bustle of the journey to hospital, but subsequent measurements are important in any patient whose condition is observed following admission. The charts may give a general clue as to the diagnosis. An increase in respiratory rate suggests pulmonary pathology rather than abdominal disease. An isolated rise in temperature certainly indicates disease but it does not specify where, nor does it necessarily signify infection. The height and the course of a fever in an adult may point to a diagnosis; in children fever is an unreliable guide as it is notoriously labile.

Consistent changes in these four vital signs over time are useful indicators of progressive pathology. A persistent rise in pulse with an accompanying fall in blood pressure is sure evidence that a peptic ulcer is still bleeding; increasing fever means that an empyema of the gallbladder needs draining. Changes in pulse and blood pressure following abdominal trauma are useful, although they usually indicate the need for active treatment rather than specifying the underlying diagnosis.

There are many signs found elsewhere in the examination which indicate disease within the abdomen. General features of the patient such as anaemia, jaundice, and facial flushing all have a direct relevance to abdominal diagnosis. The pallor of fear must not be confused with the pallor of anaemia, and cyanosis often accompanies an acute intra-abdominal catastrophe. In children, acute inflammation of the upper respiratory tract can present with abdominal pain and examination is not complete until the tonsils and the ear drums have been inspected. Here, however, we are primarily concerned with the abdominal signs.

Physical examination of the abdomen follows the time-honoured sequence of inspection, palpation, percussion, and auscultation. Many signs can be seen and few patients, even young children, object to simple observation. Palpation may be painful and it is certainly unusual. Explaining what you are doing helps a patient to relax and so does distraction with conversation. Sometimes palpation with a stethoscope is useful. Percussion and auscultation are less useful in the abdomen than in examination of the chest.

Different doctors obtain different histories and variations in the interpretation of physical signs are even more marked. Natural variation is compounded by the lack of universal agreement on the definition of some physical signs. Despite this the basic findings should be recorded in the notes. Eponymous signs are best avoided. In practice they are rarely absolutely pathognomonic of one condition.

Inspection of the abdomen is a subtle art. First and foremost both the patient and the examiner should be comfortable. The patient must lie as flat and as straight as possible with the head on a single pillow. The examiner should sit at the right hand side of the bed so that his arm and hand can lie parallel to the abdominal wall. Daylight and warmth are desirable and adequate exposure of the abdomen essential, although it is kind to keep the genitalia covered until they are actually examined.

Time should then be spent simply looking but looking in an intelligent and thoughtful way. Most important physical signs can often be seen (Fig. 2) 1421. The history will have given some clues as to possible diagnoses, and there will be specific signs to look for while remembering that negative findings are equally important. Previous abdominal operations will have been noted in the history and the scars can be examined. Their only importance now is that there may be an incisional hernia or underlying intraperitoneal adhesions. Obvious discoloration is always important. Bruising from a seat belt injury or the blue-grey discoloration in the flanks or around the umbilicus from haemorrhagic pancreatitis are both good examples.

The first thing to decide is whether the shape, symmetry, and contour of the abdominal wall are normal. Generalized distension is usually obvious except in obese patients, when it can be very difficult to decide whether the abdomen is simply fat. The most common cause of generalized distension in a woman is a fetus. Excess fluid and air in the gut and ascites are the common pathological causes of distension; this is usually symmetrical. Asymmetrical distension is best judged from the end of the bed and is caused either by a mass within the abdominal cavity or a lump in the abdominal wall. The two can be differentiated since the latter always moves with the abdominal wall whereas intraperitoneal lumps do not necessarily do so.

The abdominal wall normally moves with respiration. With the patient lying on his or her back the abdominal wall rises up on inspiration as the diaphragm descends and falls back on expiration. If this respiratory movement hurts then the patient will try to reduce or eliminate any movement by keeping the abdominal wall over the painful area still. This can often be seen and the effect can be enhanced by asking the patient to take a deep breath. Another common technique, but one that is less useful in the author's experience, is to ask the patient to blow his or her abdomen out and to suck it in. Patients with peritonitis find this painful, as they do when asked to cough. Sometimes in thin patients it is possible to see the abdominal wall muscles contract spontaneously in response to the painful stimulus. This is visible guarding.

Sometimes movement within the abdominal cavity can be seen on the surface. Aortic pulsation and fetal movements are both normal and so occasionally, in the elderly or those with gastroenteritis, is visible peristalsis. It is, however, a classic sign of intestinal obstruction. Distended loops of bowel can be seen through the abdominal wall and peristaltic contractions can often also be seen. These contractions are sometimes accompanied by borborygmi which are audible with or without a stethoscope. Patience is needed, and sometimes peristalsis can be stimulated by palpation of the abdomen.

Palpation of the abdomen requires warm hands, short fingernails, and care. By convention the doctor sits on the patient's right with the right hand flat on the abdomen in a comfortable position. Students, however, should learn to be ambidextrous because sometimes only the left side of the patient is accessible and some organs, such as the gallbladder, are occasionally easier to feel from the left.

Superficial palpation should consist of gentle movements of the whole hand. Deep palpation is achieved by gentle pressure and by flexion of the metacarpophalangeal joints whilst keeping the fingers extended. It is best to begin by asking the patient where the abdomen hurts and then to start palpating in the opposite corner of the abdomen. Work towards the painful area but do take care. Once hurt few patients will relax. The signs then become difficult to interpret and are sometimes actually misleading.

The abnormalities of importance in the acute abdomen separate into three groups. There are the signs associated with peritonitis, those which accompany a mass or enlargement of one of the solid organs, and finally those that differentiate the causes of abdominal distension.

The four signs of peritonitis are tenderness, guarding, rigidity, and rebound tenderness. Eliciting these signs is painful and it is better to see than to hear the pain. A flicker of the eyelids or a facial grimace is quite sufficient to establish the presence of pain, although guarding and rigidity are usually felt.

This is present when any palpation of the abdominal wall causes pain. It is either present or absent, although it is also possible to establish the extent of the tenderness over the abdominal wall. It is not easy to assess severity because patients vary so much in their reaction to pain. It is useful to establish where in an individual patient the pain is worst. Pain arising from the parietal peritoneum is accurately localized and patients can often point to the site of most intense pain. The examiner can also ask the patient to compare the intensity of pain by direct pressure in the four quadrants of the abdomen.

There are different opinions about the physical signs of guarding and rigidity, so the examiner must be specific about what he actually means. In the author's opinion guarding is present when there is reflex contraction of the abdominal wall muscles when the examining hand palpates the abdominal wall and thus causes slight pain. This may be seen but is more commonly felt.

Again there is no generally accepted definition of this sign but the most useful description is of an involuntary increase in the resting tone of the abdominal wall muscles. It may be localized or generalized. It is felt as an increased resistance of the abdominal wall to palpation. The intensity varies from minor increases in tension right up to the typical generalized board-like rigidity classically associated with perforation of a peptic ulcer.

This is the most important physical sign of the four. It can be a difficult sign to elicit but when present it establishes the presence of peritonitis. It occurs when inflamed visceral peritoneum moves across and irritates the parietal peritoneum and is best detected by percussion. This produces small movements of the underlying tissues, causes least pain, and can even localize the sign to specific areas within the abdomen. The classical method of detecting rebound tenderness by gross depression of the abdominal wall with the hand and then sudden release (hence the term release tenderness) is both crude and unkind and while sometimes useful should generally be abandoned. Rebound tenderness is also a symptom. Movement such as walking or the jolting of a vehicle may exacerbate the abdominal pain, and it is always worth enquiring about this whilst taking the history.

It is essential to establish the size of all the solid intra-abdominal organs during palpation and equally important to identify any abnormal masses. When the liver, spleen, and kidneys are enlarged there are certain specific signs that must be sought. When an abnormal mass is felt either within or separately from the solid organs then all the usual rules relating to the examination of lumps apply, although it may be impossible to assess swellings which lie deep within the abdominal cavity. Particular attention should be paid to the anatomical origin of the lump. Here mobility, and movement with respiration and pulsation are useful. It is always helpful to establish that a swelling is cystic. Sometimes tenderness and the other signs of peritonitis coexist with an abdominal swelling.

Abnormal abdominal distension may be caused by an abdominal swelling, but flatus, fluid, and faeces are more common. Pregnancy is generally obvious and faeces are easily discovered on rectal examination. Excess gas or fluid within the abdominal cavity is easy to demonstrate, but establishing the presence of free intraperitoneal air or ascites can be difficult.

No abdominal examination is complete without examination of the groins and the genitalia, particularly in men. Hernias are common but not always obvious. A small femoral hernia in a large woman is easily missed. If the hernia is the cause of an obstruction it will also be tense, tender, and irreducible, but it may not be very large.

Scrotal abnormalities such as testicular torsion and epididymo-orchitis can present with abdominal pain, but there are always abnormal scrotal signs on examination.

No patient likes a rectal or a vaginal examination but they are essential. Again the examination needs to be conducted with thought. Consider all the anatomical structures in the pelvis including the prostate and the cervix and look at the glove for blood or pus when the examination is finished. Rectal tenderness on the patient's right side may be the only sign of pelvic appendicitis. A swelling in a fallopian tube on vaginal examination may be the only sign of an ectopic pregnancy.

Percussion of the abdomen has three specific uses. First, it is the best method of eliciting rebound tenderness. Second, it is the most sensitive method for detecting enlargement of the bladder. Third, shifting dullness determines the presence or absence of ascites. It has a subsidiary role in confirming the size of the liver and spleen and may sometimes be useful in outlining an intra-abdominal mass.

Auscultation of the abdomen is not very helpful, but the presence or absence of bowel sounds is a useful physical sign. Qualitative observations are less reliable. Nevertheless an increase in the magnitude and the frequency of bowel sounds accompanies mechanical intestinal obstruction whilst a succussion splash, which can sometimes be heard without a stethoscope, is a sure sign of obstruction. Bowel sounds which definitely disappear during observation of a patient with abdominal pain and tenderness indicates the onset of peritonitis and the need for a laparotomy.

Although investigations are more or less routinely requested in most patients with acute abdominal pain, very few of the tests are actually valuable in making a diagnosis. In a few patients no investigations are necessary because the diagnosis is clinically obvious. In the majority the cause of the pain is initially uncertain and it is hoped that tests will help. Older and more experienced surgeons maintain that it is preferable to wait and see in these circumstances. They argue that significant disease is usually progressive and when the patient is re-examined after an interval the physical signs are more marked and the diagnosis easier. Younger surgeons think that the delay gives time for complications to develop with a consequent increase in postoperative morbidity that diagnostic investigations might avoid. However, their enthusiasm for investigation can also delay a necessary operation if the tests take too long to perform. In a few patients an accurate diagnosis which is essential for correct treatment can only be made with the help of special investigations.

We are most concerned here with tests that will help in the diagnosis and the management of the patient within the first 24 h of admission. After that the number of tests that can sometimes be useful is vast and they are considered in the individual subject chapters. Analysis of venous blood and various radiographs are the most popular immediate investigations, with the recent addition of ultrasound examination. They can be divided into two groups—those tests that help in diagnosis and those that help in management.

Simple clinical inspection of the urine should still be regarded as an essential part of examination of the abdomen. Urine containing tiny amounts of blood looks smoky. Infected urine smells unpleasant and may be cloudy or even contain frank blood. Sugar and ketone can also both be smelt and confirmation of all these findings using biochemical sticks is convenient and easy. Most clinicians should be able to identify pus cells in centrifuged urine using the ward microscope.

When a urinary tract infection is suspected a carefully collected urine specimen should also be sent immediately to the laboratory for analysis. It is not easy for any patient to provide a true mid-stream urine specimen, and they must be both helped and supervised. Even then contamination can be a problem and there are occasions, particularly in women and children, when a catheter specimen should be collected. Urethral catheterization is usually appropriate but suprapubic puncture of the bladder provides the least contaminated specimen and carries the least risk of introducing an infection. Even though the culture result will not be available for a few days the sample must be sent acutely otherwise the opportunity to identify the organism responsible may be lost, as most patients with a urinary tract infection presenting with acute abdominal pain will need immediate treatment with antibiotics.

Many significant causes of acute abdominal pain are associated with some degree of inflammation. As a consequence an increase both in the absolute numbers of white cells and in the proportion of neutrophils might be expected. The reverse observation is also true: an increase in the white cell count indicates the existence of inflammation. It is always necessary to interpret the result in the clinical context, for the inflammation may not necessarily lie within the abdomen. A value within the normal range does not exclude intra-abdominal inflammation.

This very simple way of looking at the white cell count is not the most useful. It is more helpful to interpret the result in a statistical sense. In other words the probability of a patient with a normal white cell count having acute appendicitis, for example, is low whilst the chances of appendicitis in a patient with a raised count are higher (Table 2) 400.

The same observations may be made about an excess of neutrophils in the differential white cell count. Indeed the results of all such tests used to establish a diagnosis should ideally be analysed in this way. In practice a normal white cell count is often used to reassure the surgeon who wants to wait and see while an increased count supports a decision to operate. The surgeon should realize, however, that the test is then being used to help in a management decision and not to make a diagnosis.

Acute pancreatitis usually presents with the symptoms and signs of peritonitis, and normally patients with peritonitis warrant an immediate laparotomy. Surgery is, however, best avoided in patients with acute pancreatitis. The rise in serum amylase concentration which usually accompanies pancreatitis allows the correct diagnosis to be made and a laparotomy is thus averted.

Because the result is so important for both diagnosis and treatment it is essential to appreciate the limitations of the test. Other intra-abdominal catastrophes such as a perforated peptic ulcer, a ruptured aortic aneurysm, or dead gut can cause a modest rise in the serum amylase level, while if the blood sample is taken too long after the onset of the pancreatitis the level may have reverted to normal and so give a false-negative result. Again a statistical approach can be adopted. A low level of serum amylase carries a low chance that the patient has acute pancreatitis whilst a high level implies a high chance (but not a certainty) that pancreatitis is indeed the diagnosis.

Controversy surrounds the use of plain abdominal radiology. Sometimes the films confirm the clinical diagnosis, add further detail, and modify the management of an individual patient. At other times the films are simply misleading, although occasionally they suggest a diagnosis which the clinician has not considered. One thing is certain. Not every patient with acute abdominal pain needs an abdominal radiograph. When it is requested the doctor should be clear what information he hopes to gain and he must have the skills to interpret the films if no radiologist is available.

Traditionally two films are taken, one with the patient lying supine and the other with the patient sitting or standing erect. Modern protagonists of a single supine film point out that little additional information is derived from the erect film and add that not every patient with acute abdominal pain can safely or comfortably sit or stand. Some radiologists prefer, as an alternative to the erect film, to lie the patient on their right side and then take a lateral radiograph (the lateral decubitus view). In the author's opinion an erect view does, on occasion, add useful information whereas the lateral decubitus view usually does not. It provides only a limited view of the abdominal cavity and free intraperitoneal gas is better seen on a chest radiograph.

Abdominal films are more use in some circumstances than in others. None of the radiological signs of acute appendicitis are truly helpful, but radiological examination should be performed in patients with suspected intestinal obstruction and those who have suffered abdominal trauma (Fig. 3) 1422. Stones in the kidney, the ureter, or the gallbladder are sometimes confirmed on a plain film, and calcification of the wall of an abdominal aortic aneurysm may be the only clue to its presence. Radiology of the abdomen is more useful in older patients, who tend to have more significant pathology and thus more abnormalities on such films. It is important to remember that the presence of abnormalities on any abdominal radiograph is valuable, but their absence is meaningless.

A good quality erect chest radiograph is the best film with which to confirm the presence of free intraperitoneal air (Fig. 4) 1423. This can be seen as a black crescent, sometimes with an air–fluid level, underneath one or both diaphragms. Proximal perforations of the bowel tend to lead to larger amounts of free air than distal perforations; if the perforation has occurred some time before presentation, as can happen in patients with diverticular disease, the margin of the pneumoperitoneum is often rather hazy and irregular. There may also be a small pleural reaction above the diaphragm.

In very old and very young patients pneumonia and pleurisy may present with referred abdominal pain. Lower rib fractures may indicate a ruptured spleen or lacerated liver.

Renal colic is usually an easy clinical diagnosis to make because of the characteristic distribution of the pain. When the diagnosis is in doubt an emergency intravenous urogram is frequently helpful: delayed excretion of contrast from the kidney on the side affected by pain confirms the diagnosis. A normal urogram effectively excludes the diagnosis, provided the examination is done during or within a short time of an episode of pain. Other causes of abdominal pain can then be considered. Although the film taken immediately after injection of contrast is sufficient for diagnosis, the examination is usually completed in order to determine the site of any obstruction, the size of an offending stone, and the degree of dilatation of the system.

Intravenous urography is also useful in patients who have suffered trauma to the urinary tract. Most such patients have haematuria. The degree and the site of any damage may be displayed, and the presence of a normally functioning kidney on the unaffected side can be confirmed.

Clinical examination of the abdomen in the trauma victim is notoriously unreliable and computed tomography (CT) of the abdomen as a means of diagnosis is rapidly becoming routine in the seriously injured patient. Ultrasound is more readily available and should probably be the first investigation, but the examination is more limited.

Apart from trauma there are few indications for the use of CT in the immediate diagnosis of abdominal pain. It may be of occasional help in the diagnosis of an aortic aneurysm and acute pancreatitis, although in the latter case CT is of greater value in the identification of complications.

Ultrasound is widely used in the diagnosis of acute abdominal pain. Its place in elective diagnosis of conditions affecting the upper abdomen, the pelvis, and the retroperitoneum is already established; it is also useful in the emergency patient. Gallstones and an aortic aneurysm are easily seen (Fig. 5) 1424, as are the oedematous gallbladder wall and a tear in an aneurysm. Hydronephrosis, and sometimes stones in the kidneys or ureter, can be seen, and ovarian cysts and swellings on the fallopian tubes can be identified in the pelvis. The ultrasound probe can also be used, like the examining hand, to identify the specific structure that hurts.

Ultrasound is less useful in examining the bowel because of the presence of gas. However, the inflamed appendix often lies behind the caecum and contains little air. Certainly the ultrasound probe can localize the tenderness to this specific area and can sometimes demonstrate an oedematous tubular structure at the site where a retrocaecal appendix should lie.

Following trauma, ultrasound can demonstrate the presence of free intraperitoneal fluid and identify damage to the liver, spleen, kidneys, and pancreas. It cannot identify blood clot very well and it is of no practical use in looking for injury to the gut.

Doppler ultrasound, which demonstrates flow in vessels, can help decide the cause of acute testicular pain. The hyperaemia of epididymo-orchitis is in marked contrast to the ischaemia of torsion.

Many of the tests that are useful in diagnosis also have a role in management. A progressive reduction in the white cell count and improvement in the radiological signs of obstruction after treatment both indicate resolution of the pathology. A large number of other tests also help in the treatment of a patient, many of which are undertaken soon after the patient is admitted to hospital. Some of them also play a part in diagnosis.

The clinical diagnosis of anaemia is not always reliable, and in any patient who may possibly have an anaesthetic it is clearly important to know the oxygen-carrying capacity of the blood. The initial haemoglobin value does not indicate the volume of blood lost in patients with overt evidence of acute haemorrhage, but sequential measurements can give a rough guide, provided any blood transfused is taken into account. Occasionally the discovery of an unexpectedly low haemoglobin level can help in diagnosis: carcinoma of the caecum as a cause for intestinal obstruction with anaemia is a classic example. Packed cell volume accurately reflects the severity of fluid loss in a dehydrated patient and it is a good guide to the adequacy of rehydration.

Most patients with major intra-abdominal pathology should have serum creatinine and electrolyte concentrations measured on admission. The initial values must be interpreted in the clinical context, particularly if the patient is dehydrated; in most circumstances it is the serum potassium concentration which is the most important because of its role in cardiac function. Serial values are vital for proper postoperative fluid management.

Most patients with an acute abdomen due to liver and biliary disease are jaundiced. The depth of the jaundice reflects the severity of the pathology, and it is rare to need to measure the liver function tests acutely. It is, however, essential to obtain a blood sample on admission for later analysis because subsequent deterioration in biochemical parameters, which may not be clinically obvious, will demand further action. This particularly applies to the elderly in whom the signs and symptoms of biliary disease are often obscure. The diagnosis is sometimes not even considered until abnormal liver function tests are discovered.

This is only of immediate value in patients with acute pancreatitis. Depleted values are an indirect guide to the diagnosis and are used in some severity scoring systems. When low calcium levels threaten to induce tetany, intravenous calcium supplements are needed. Calcium level is always measured in patients with renal colic, in whom evidence of hyperparathyroidism is sought, but hypercalcaemia is rarely found.

Analysis of an arterial blood sample should be performed in a patient who is severely ill with an acute abdomen from whatever cause. Many such patients are covertly hypoxic, and the result of blood gas analysis may indicate the need for immediate ventilatory support. More commonly, patients will need ventilation after an emergency operation; preoperative values are then a useful indicator of the patient's progress.

Blood gas analysis is also a component of many scoring systems to assess the severity of acute pancreatitis.

If a chest radiograph is not necessary for diagnosis it is unlikely that it will be needed in the management of the patient. Nevertheless there are times when, although a clinical diagnosis can be made, a chest radiograph should be obtained simply to provide a baseline. It is often useful to know that postoperative changes in a number of parameters, particularly the chest radiograph, were not present before surgery.

It is unusual for contrast radiology to be performed as an emergency although an urgent barium or air enema should be undertaken in a child with suspected intussusception. In adults with large bowel obstruction a limited barium enema examination is sometimes useful to establish the presence of a mechanical obstruction rather than pseudo-obstruction. In patients with small bowel obstruction where the cause is obscure or resolution is not occurring as fast as expected a small bowel enema is always useful.

Anyone over the age of 40 who presents with acute abdominal pain, particularly if the diagnosis is not straightforward, should undergo electrocardiography. Very occasionally a myocardial infarct will present with abdominal pain and recovery is unlikely to be helped by an unnecessary laparotomy.

Emergency gastroscopy and colonoscopy, occasionally performed on the operating table, are helpful in patients who present with acute gastrointestinal haemorrhage. Precise localization of the bleeding point is essential for effective treatment. Mesenteric angiography may also be needed. In both instances treatment as well as diagnosis may be possible.

Endoscopy has no part to play in the diagnosis of a perforated peptic ulcer, and may make matters worse by blowing air into the peritoneal cavity through the perforation.

This is a useful investigation in patients with abdominal trauma, particularly if they are unconscious or are otherwise unable to co-operate in an abdominal examination. The presence of significant amounts of blood or intestinal contents in the washout fluid is a clear indication for an urgent laparotomy.

In patients with acute abdominal pain the presence of excess neutrophils in fluid aspirated from or washed out of the peritoneum is a reliable indicator of peritonitis. The test is not widely used, probably because it does not indicate the underlying cause of the inflammation.

General surgeons have been slow to use the laparoscope, despite very good evidence that it can help in making a diagnosis. All that is set to change, partly because of the introduction of new video technology but also because of the development of laparoscopic surgery.

Many surgeons will consider that the opportunity to remove an appendix at the same time as confirming the diagnosis justifies the slight risks that laparoscopy entails. The current technique for laparoscopic appendicectomy is time-consuming and cumbersome, but McBurney's incision will undoubtedly become a relic of the past.

Laparoscopy also has a role in the management of the abdominal trauma victim. However, it is only used as a more sophisticated form of peritoneal lavage. If blood or intestinal contents are found in the peritoneum then the endoscopist must look for the source and also decide if the severity of the bleeding is sufficient to justify a laparotomy.

No one really understands how a doctor makes a diagnosis, although the process has been analysed many times. In theory it is simply a matter of collecting all the relevant facts and analysing them correctly. The contrast between the junior clinician who takes time and trouble over the patient and yet makes the wrong diagnosis half the time and his senior colleague who asks a few questions, performs a limited examination, and is right eight times out of ten shows that this is not the whole story. Very few patients present with all the symptoms and signs of their disease and only experience can teach the clinician which few questions to ask, how to ask them, and how to interpret the answers correctly in the context of the individual patient. The last skill is particularly important when some of the facts conflict. Experience and constant practice are certainly essential for maximum accuracy.

In actual clinical practice, several methods are used to make a diagnosis. Some involve purely practical considerations whilst others look at the same data in different ways. Most clinicians use all the methods at one time or another, often together, and usually without giving the matter a second thought.

In this unusual circumstance the patient gives the typical history of a classical cause of abdominal pain with every symptom in its correct place. Examination reveals all the expected signs and the diagnosis is obvious even to the least experienced clinician. There is really no place for a differential diagnosis nor are investigations necessary. All that is left to be done is to organize the correct treatment.

More commonly a variety of diagnoses is suggested by the patient's initial complaint. Most doctors then ask further questions to support or exclude each individual diagnosis. The method reaches its peak with the most experienced clinician who makes a diagnosis on the basis of half a dozen questions and a very limited examination of the abdomen. This is the culmination of a natural development in most doctors. To begin with they are taught to obtain information in a systematic way and to record it all in a standard format. Later they learn that certain symptoms and signs are commonly associated with certain conditions and so they construct in their minds algorithms for each individual patient. With experience the method works well but it is dangerous for the beginner because of the risk of following the wrong pathway early in the decision tree.

Another approach is to consider which anatomical structure could be the cause of the symptoms and signs. Abdominal anatomy is broadly the same in everyone and most patients can localize their pain to one area of the abdomen. Each organ within that area is then considered in turn as the source of the symptoms. The most difficult area is the left upper quadrant where there are no structures which commonly cause acute abdominal pain. Anatomy can also be used to construct a list of all the other structures that could cause abdominal pain at any particular site.

A particular pathological process is suggested by certain symptoms and signs. This may, in turn, suggest a diagnosis. Acute intra-abdominal inflammation is usually accompanied by a fever and abdominal tenderness, while mechanical obstruction is characterized by colic and vomiting. Sometimes obstruction and inflammation coexist, but this in itself shortens the diagnostic list to only a few possibilities.

The importance of sex in relation to the possible causes of abdominal pain is obvious but age is an equally valuable discriminator. Cancer is more common in the old than the young, for example, whereas non-specific abdominal pain is a disease of the young.

Acute appendicitis and non-specific abdominal pain account for approximately two-thirds of all the patients admitted to hospital with abdominal pain in Europe and North America (Table 1) 399. This means that for two-thirds of the time the only important distinction the surgeon has to make is between these two conditions. Looked at another way this also means that if it was possible to separate these two diagnoses accurately from all the other possibilities then the clinician would make a correct diagnosis in two-thirds of the patients.

Common things occur commonly. The most common surgical cause of acute abdominal pain in a patient admitted to a hospital anywhere in the world is acute appendicitis. The next most common cause in Africa is small bowel obstruction; in the West it is acute cholecystitis. Knowledge of local epidemiology is therefore useful in the diagnosis of abdominal pain, although the patterns of disease will change over time.

This is essential for the young clinician but a luxury for the surgeon in charge who has to decide on treatment. The latter keeps a list in his head but the former is wise to write the possibilities down, preferably in order of probability. There is truth in the saying that you will not make a diagnosis unless you think of it: making a list may jog the memory, as the rarities are easily forgotten. However, a list of possible diagnoses only provides a framework on which to base further enquiry. It is of no direct help in planning treatment.

Most patients with abdominal pain will undergo some investigations on admission, particularly if the diagnosis is not obvious. The results usually support or refute a diagnosis that has already been considered, but they sometimes suggest another condition. This applies particularly to abdominal radiographs. It is then always worth repeating the clinical examination at once to test whether this possible diagnosis would fit with the findings. It is true that this is the reverse way to interpret information but it is, nevertheless, often a good way to make a correct diagnosis.

Sometimes the initial symptoms and signs are insufficient to make a diagnosis. In these circumstances time is an invaluable ally. If spontaneous improvement occurs over the ensuing few hours then the need to make a prompt diagnosis lessens; if the disease process progresses the symptoms and particularly the signs will worsen. In practice the clinician returns to examine the patient again 3 or 4 h after admission. If the symptoms and signs are worse the diagnosis will usually be apparent. If there is improvement, no surgical treatment is needed, but it is also probable that no specific diagnosis will ever be made. If there has been no change, a further examination a few hours later is needed. Some people call this approach ‘active observation’.

There are two risks attached to a ‘wait and see’ policy. The first is that the passage of time may allow a complication to develop with a consequent increase in morbidity. The second is a lack of intellectual rigour: the clinician waits until the patient has recovered or until it becomes obvious that a laparotomy is essential and never bothers to make a clinical diagnosis.

Only a clinician can obtain the information that is needed from a patient with acute abdominal pain and decide on the best form of management. This information has first to be analysed in order to make a diagnosis before treatment can begin. It is here that computers can help, although this help does not derive entirely from sophisticated data analysis. In the Western world the diagnostic accuracy of the first doctor who sees a patient with acute abdominal pain in hospital is about 45 per cent. Improving the quality of the information obtained by the rigid definition of symptoms and signs and the amount of information collected with the use of special forms increases this initial accuracy to 60 per cent. This figure can be improved to about 70 per cent by giving feedback about their accuracy to individual doctors and by computer analysis of the data from individual patients in comparison with a large database of information from patients with abdominal pain of known cause. When the results of investigations are available, accuracy further improves to 75 per cent. Performance is improved at every level of surgical experience, and the very best clinicians can make an accurate diagnosis of the cause of acute abdominal pain 80 per cent of the time. This improvement in diagnostic accuracy is also reflected in an improved outcome for the patient, with a substantial reduction in the number of normal or perforated appendices removed at operation and fewer serious surgical errors.

The pragmatists claim that the diagnosis lies between acute appendicitis and non-specific abdominal pain in two-thirds of the patients; the former need an operation whereas the latter do not. They also claim that many of the remaining one-third of patients with another diagnosis will benefit from a laparotomy when the signs indicate that one is needed. The clinician faced with a patient with acute abdominal pain therefore has only one decision to make. Do the symptoms and signs justify a laparotomy? If the answer is yes then the diagnosis will become obvious at operation. If the answer is no then a period of active observation is all that is needed. The main disadvantages are a high negative laparotomy rate and the complications that ensue from inappropriate operations.

There are occasions when a practical approach has to be adopted. It may be obvious that a laparotomy is essential, but a diagnosis cannot be made because the patient is unconscious or is too ill to co-operate. Alternatively there may be no time to make a complete diagnosis. This will certainly be the case when the available surgical services are overwhelmed by multiple casualties. Absent bowel sounds as an indication of peritonitis would often be sufficient justification for a laparotomy in these circumstances.

It is a good rule that patients with acute abdominal pain are not given analgesia until a diagnosis has been made and treatment planned. However, common humanity often demands that pain is relieved before a surgical opinion is obtained. Provided that the time of administration and the dose of the drug is recorded, an experienced clinician should not be deceived. Analgesia does not eliminate all the physical signs but it does subdue them and due allowance must be made. Once a diagnosis has been made every patient must be given adequate analgesia at once, whatever other treatment is planned.

Some patients cannot be properly examined because of the severity of their pain. Adequate analgesia, perhaps given intravenously, may then make an examination, and a more accurate diagnosis, possible.

This is included only to be condemned. Immediately visible symptoms and signs in the abdomen can often be interpreted in several ways. Furthermore, the doctor who jumps to conclusions tends to make subsequent observations fit his chosen diagnosis rather than to analyse them dispassionately. There are no circumstances where an instant diagnosis of the cause of an acute abdomen should be made. It is simply not safe.

How does the young clinician obtain the skills to match those of his seniors in the diagnosis of acute abdominal pain? A determination to make the correct diagnosis on every occasion and plenty of experience are the basic answers but there are a few techniques that will help.

Computer analysis has shown the importance of obtaining all, but no more than all, the necessary information to make a diagnosis. Whilst it is usual for the house surgeon or intern to record a complete case history, residents or registrars should confine themselves to recording in their own writing the facts relevant to the diagnosis. This forces the clinician to decide which facts are important and starts the process of excluding more information than can be processed. This is easier if a special data collection form is used.

The surgical trainee must always decide upon and write down one diagnosis even if a list of other alternatives is appended. After all, a single diagnosis will become essential as soon as the trainee takes on the responsibility for deciding on treatment. To begin with the diagnosis will be wrong at least half the time, but as confidence and experience is gained this figure will improve.

Asking a senior colleague for advice about the diagnosis and management of a patient is always good training. The discipline of putting the right facts together in a logical order and presenting them correctly is excellent practice at analysing the information.

It is always worth comparing the findings on admission with the final diagnosis. It is often possible to identify where a sign or symptom was misinterpreted and also to appreciate that certain groups of findings correlate with certain diagnoses.

It is not the intention here to discuss in detail all the possible causes of acute abdominal pain. Rather it is to point out certain features which will help in the differentiation of the more common causes of pain and also to discuss a number of unusual problems which are not dealt with elsewhere.

There are various ways in which to classify the causes of acute abdominal pain. Bailey and Love compared the differential diagnosis of acute appendicitis to a house with two storeys corresponding to the upper and lower abdomen, the pelvis (‘the basement’), the thorax (‘the attic’), and the retroperitoneal structures (‘the backyard’) (Table 3) 401. The underlying anatomical analysis is still useful, but in practice many experienced clinicians will first decide whether the features of the case suggest inflammation, obstruction, colic, a major catastrophe, or simply the presence of a mass. In many instances more than one pathology is involved. Inflammation can cause obstruction and unresolved obstruction will eventually lead to inflammation. Nevertheless the distinction is useful in practice because within each category there are then a number of possible causes for the findings.

Inflammation within the peritoneal cavity usually arises from one of the intra-abdominal organs although primary peritonitis does rarely occur in young girls. In most instances the inflammation is secondary to infection or ischaemia. Initially the inflammation is confined to the organ of origin and this makes the diagnosis easier as the physical signs will also be localized. Untreated and progressive inflammation will eventually lead to gangrene and necrosis, with consequent perforation of the viscus and the development of generalized peritonitis. Both the location of the initial signs and the speed of onset of generalized peritonitis will give a clue to the original site of the inflammation. The prognosis worsens with the progression of the pathology and so the clinician is always anxious to treat localized peritonitis before perforation occurs.

Many clinicians consider this to be a complete diagnosis in itself but in older textbooks acute catarrhal appendicitis, acute obstructive appendicitis, acute perforated appendicitis, and an appendix mass are all described as separate clinical entities. The last two certainly present a different clinical picture from the first two, and sometimes it is even possible to identify the colic and the vomiting which marks the start of an obstructed appendicitis. But with the exception of an appendix mass, where conservative management is usually preferred, the other three varieties all require an early operation so that differentiating between them is hardly worthwhile. Even so it is always wise to insist on adequate resuscitation of the patient with perforated appendicitis before surgery.

The other common error is to assume that all the patients with acute appendicitis present to a surgeon and, furthermore, that they all undergo appendicectomy. Acute inflammation sometimes resolves spontaneously, and acute appendicitis is no exception. A proportion of patients diagnosed as having non-specific abdominal pain probably have mild acute appendicitis, and a few such patients are readmitted a few days or weeks later with clear-cut appendicitis or an appendix mass.

Half the patients manifest all the classical symptoms and signs of appendicitis but the other half present in a variety of ways which varies from the misleading to the bizarre. The diagnosis is often difficult at the extremes of life and during pregnancy. The young may not be keen to reveal their symptoms and in the old there are often few physical signs. Stoical and muscular young men may have convincing symptoms but trivial tenderness. Pregnancy distorts the anatomy and the uterus can itself be the source of the pain. Diarrhoea and vomiting with abdominal pain will suggest gastroenteritis to most doctors, but they can be caused by appendicitis; appendicitis may also accompany gastroenteritis. The variable anatomical position of the appendix in normal people can also mislead the clinician. There may be few abdominal signs with pelvic appendicitis and the doctor who omits the rectal examination will miss the diagnosis. Retrocaecal appendicitis often causes microscopic haematuria due to inflammation of the adjacent ureter, and this can lead to an incorrect diagnosis of urinary tract infection or ureteric colic.

Bizarre presentations include a lumbar abscess from missed retrocaecal appendicitis, a perineal sinus from pelvic appendicitis, and appendicitis in the presence of an appendicectomy scar.

The list of differential diagnoses is very long: many conditions can mimic appendicitis, from right-sided pneumonia to salpingitis, and including a ruptured aortic aneurysm (Table 3) 401.

At least one in five appendices removed at operation is normal on histology. This seldom matters, for another surgical cause for the pain is found at operation in about half these patients, although the complications that can ensue from an unnecessary operation should not be overlooked. A greater danger is to delay an operation until perforation has occurred, when the patient is worse, surgery is more difficult, and the complication rate is higher. The surgeon who never removes a normal appendix is undoubtedly exposing other patients to the risk of perforation.

Most patients with abdominal pain never discover the cause. Only about one patient in every 15 with pain is admitted to a hospital and of those who are admitted four out of every 10 leave without a diagnosis. Some patients find these statistics reassuring, but others find the doctor's inability to find a cause for their symptoms distressing. It is particularly important to reassure the latter group that a diagnosis of non-specific abdominal pain does not imply that there is no pain nor that there is no cause. The alternative title of non-surgical abdominal pain is more accurate in the sense that these patients do not need an operation, but it is inaccurate since there may still be an underlying surgical problem.

The predominant symptom is always pain, but there is usually abdominal tenderness as well, often in the right iliac fossa. This normally implies peritonitis but the signs are never sufficient to justify the diagnosis. Some people call this peritonism implying thereby irritation of the peritoneum without inflammation; this term is best avoided as it has no true pathological explanation.

Such patients are usually actively observed and the symptoms and the signs subside as mysteriously as they came. Up to a point, therefore, it is a diagnosis of exclusion and one that is only made in retrospect, but there are a few pointers towards a positive diagnosis. Most of the patients are young and two-thirds of them are women. The pain is rarely made worse by movement, it rarely moves its position in the abdomen, and about one-third of patients will keep their eyes closed during the abdominal examination (the ‘closed eyes’ sign).

Some of these patients do in fact have minor versions of recognized clinical conditions such as mesenteric adenitis, threadworm infestation, gynaecological pain from ovulation, or torsion of a colonic appendix epiploicae. Incomplete intestinal obstruction may not be clinically obvious and it may resolve before the diagnosis is made if the loop of bowel releases itself or the adhesion tears. Obscure abdominal pain in the elderly, which is uncommon, is often associated with cancer, particularly of the colon. Social and psychological factors play a very important role in some patients. This is usually because of anxiety about the minor abdominal pains which afflict everyone at some time or another rather than being a primary cause in themselves.

This is usually an easy diagnosis to make and even easier to confirm with the use of ultrasound. In the developed world it is the most common cause of acute abdominal pain in elderly people and the third most common cause in the younger age groups.

There are three important circumstances in which the diagnosis is difficult. In the elderly the symptoms are sometimes obscure and the signs minimal. There is often no fever and the white cell count is normal. Acute acalculous cholecystitis can develop insidiously in patients who are very ill for another reason. They are often being fed intravenously in intensive care. Ultrasound can confirm the diagnosis but only if it is considered. The third difficulty arises when a patient gives a classical history of acute gallstone disease but has no stones on ultrasound examination, even though the ultrasound probe may identify the gallbladder as the source of the pain. Sometimes the gallstones are too small to be seen, and occasionally the patient has passed their only stone. The diagnosis is then only made in retrospect when the stones have either reformed or grown in size. Alternative explanations for acute right upper quadrant abdominal pain in the absence of gallstones are exacerbation of a peptic ulcer, renal colic, and chlamydial perihepatitis (the Curtis-Fitzhugh syndrome).

Sometimes acute pancreatitis can be confidently diagnosed on the basis of the patient's symptoms and signs. A raised serum amylase merely confirms the clinical opinion. Constant epigastric pain radiating to the back and bruising in the flanks or around the umbilicus, which can be very difficult to see, are the key features. At other times it is difficult to distinguish pancreatitis from other causes of upper abdominal peritonitis, particularly a perforated peptic ulcer. The problem is compounded because a mild elevation of serum amylase is sometimes seen with a perforation and also in patients with ischaemic bowel or acute cholecystitis. Acute cholecystitis and acute pancreatitis can occur together. In the acute management of the patient with pancreatitis it is rarely helpful to decide whether gallstones or excess alcohol are the cause. Inflammation tends to be less severe in patients when the acute episode is superimposed on chronic pancreatitis.

Pancreatitis can also suddenly develop after any abdominal operation, and it is then almost impossible to diagnose clinically. It presents simply as an unexplained failure to recover from the surgery, perhaps accompanied by vomiting for no apparent cause. The diagnosis is only made when the serum amylase is measured as part of a biochemical screen.

This reinforces the general point that it is always wise to measure the serum amylase in any patient with predominantly upper abdominal peritonitis and then to re-evaluate the patient in the light of the result. High serum levels generally support the diagnosis of acute pancreatitis; modest elevations demand consideration of other possibilities.

This may present in a variety of ways, the most common of which is acute localized inflammation of a diverticulum. Less common presentations are generalized peritonitis from perforation of a diverticulum, a pericolic abscess, intestinal obstruction from adhesions, a faecal fistula, and acute rectal haemorrhage.

Acute pain and tenderness in the lower abdomen are the hallmarks of acute diverticulitis. Initially the symptoms and signs are fairly widely spread, even into the upper abdomen, but as the inflammation resolves the signs tend to localize to the left iliac fossa. There is usually an alteration in bowel habit and often frequency of micturition. A rectal examination is essential as tenderness or even a mass may be palpable; the most important differential diagnosis is carcinoma of the large bowel. A barium enema or a colonoscopy is best deferred until the acute episode has subsided because of the risk of perforating the inflamed diverticulum.

Some patients operated upon for appendicitis have acute inflammation of the terminal ileum. Crohn's disease can certainly present in this way, although there is usually a history of diarrhoea and weight loss as well as the recent pain. Infection with Yersinia enterocolitica can also cause a self-limiting ileitis. The diagnosis is made by sequential serological studies.

This is a rare condition which presents in a similar fashion to acute appendicitis. At operation the appendix is normal and a large mass is felt in the caecum or ascending colon. Most surgeons mistake this mass for carcinoma and so perform a right hemicolectomy. If the correct diagnosis is made resection is unnecessary.

Rupture of an ovarian follicle, ectopic pregnancy, salpingitis, ovarian cysts, and fibroids are common gynaecological causes of acute abdominal pain. Most of them present with other abdominal symptoms and signs which lead to confusion and a mistaken diagnosis of acute appendicitis, but all of them can be diagnosed clinically. A vaginal examination is essential. Reliable early pregnancy tests and pelvic ultrasound are also valuable aids to diagnosis, as is laparoscopy.

Salpingitis is the most common gynaecological cause of lower abdominal pain in women. It is often bilateral, is often accompanied by a fever and a vaginal discharge, may be associated with an intrauterine contraceptive device, and tends to be persistent and recurrent. The normal rupture of an ovarian follicle in the middle of the menstrual cycle (mittelschmerz) is sometimes accompanied by sufficient bleeding to cause significant lower abdominal pain. The diagnosis can often be made on ultrasound by the presence of a small amount of fluid in the pouch of Douglas. The bleeding is rarely of any magnitude and no treatment apart from rest and analgesia is usually needed. Haemorrhage from an ectopic pregnancy is often frightening and can be lethal. The difficulty is that many patients do not even realize that they are pregnant. Tactful but specific questions about the possibility of pregnancy are essential: never forget that previous sterilization does not mean that further pregnancy is impossible. Clips come adrift, ties come undone, and tubes recanalize. The residual damage to the fallopian tube means that an ectopic pregnancy is more likely than usual.

Benign ovarian cysts twist; malignant cysts can rupture, bleed, infiltrate surrounding structures, and cause small bowel obstruction. Twisted cysts are easy to diagnose as there is a tender central lower abdominal mass arising out of the pelvis, although they are not always easy to distinguish from a degenerating fibroid.

This can mean anything from severe acute pyelonephritis to mild cystitis. The symptoms and the signs vary accordingly. Mild cystitis rarely presents to a hospital, but pyelonephritis certainly does and right-sided infection is easy to confuse with acute appendicitis. Normally the urine is obviously infected on testing but occasionally acute inflammation of the renal parenchyma can precede urinary symptoms and the appearance of pus in the urine by a day or two.

A perirenal abscess and a pyonephrosis both give rise to abdominal pain, but there are usually many other physical signs to suggest the true diagnosis.

Occasionally patients with testicular torsion present with lower abdominal discomfort as the predominant symptom rather than severe scrotal pain. Confusion only arises when examination of the genitalia is omitted as part of the abdominal examination. The symptoms of acute epididymitis and torsion of a testicular appendage usually focus on the scrotum rather than any abdominal pain.

It is an exceptional diagnostician who can separate Meckel's diverticulitis from acute appendicitis. It can be done because the abdominal tenderness lies closer to the midline than with classical appendicitis but there are no other distinguishing features. The difference is of no practical importance since the true diagnosis is soon apparent at operation.

Generalized peritonitis is easy to diagnose, but the causes are legion. Up to a point identifying the precise cause is irrelevant, because a laparotomy is mandatory if the patient is to survive. The diagnosis is then immediately apparent. However, any surgeon knows the difficulty of oversewing a perforated duodenal ulcer in a fat patient from a lower left paramedian incision, and one incision is better than two for the patient, leaving aside the benefits to the surgeon's pride. Any abdominal organ can rupture as a result of inflammation, ischaemia, or trauma and flood the peritoneum with blood, bile, urine, or intestinal contents. It is always worth trying to decide which fluid is present and from whence it arises.

Perforation of the gut with leakage of intestinal contents is the most common problem. Bowel content is the most irritant substance and gives rise to all the classical symptoms and signs of peritonitis. The three common causes are a perforated peptic ulcer, perforated diverticular disease, and perforated appendicitis. Stercoral perforation of the colon due to severe constipation is a rare variant of perforated diverticulitis. Distinguishing between the various causes depends on taking a very careful history to try and decide the symptoms at the onset of the illness and an equally careful evaluation of the abdominal signs. Some patients are simply too ill to co-operate. Patients with a perforated ulcer may give a history of indigestion or the consumption of ulcerogenic drugs. Sometimes it is possible to decide that the signs are worst in one particular area of the abdomen.

Blood, bile, and urine are all rather less irritant to the peritoneum and give rise to less marked physical signs. Blood, which usually comes from a ruptured spleen, an ectopic pregnancy, or a ruptured aortic aneurysm characteristically gives rise to tenderness and rebound tenderness but very little guarding or rigidity. Uninfected urine and bile can both be present in the peritoneum in large amounts with very few signs at all, although once infection is present the signs are usually very marked.

Abdominal pain, vomiting, abdominal distension, and constipation are a quartet of very obvious symptoms and signs which make the diagnosis of intestinal obstruction easy. If the patient's presentation to the doctor is delayed the hyperactivity of the bowel which is so obvious at the onset of the obstruction may have been replaced by paralysis, with disappearance of the colicky pain and any visible peristalsis.

Not every patient with obstruction requires immediate surgery; it is important to try and decide on the cause. Uncomplicated obstruction due to adhesions, which is now the most common problem, is best treated conservatively to begin with, as many episodes will subside spontaneously. An obstructed hernia, which means occlusion of the lumen of the bowel within the hernia, requires a prompt operation. Strangulation implies impairment of the blood supply to the bowel and also demands an urgent operation. It is indicated by a shocked patient with peritonitis as well as the signs of obstruction, although these will not be present if only omentum or extraperitoneal fat are trapped in the hernia. An inguinal or incisional hernia should be obvious, but it is extraordinarily easy to overlook a small tense femoral hernia in a fat patient. The various rare intra-abdominal hernias are rarely diagnosed before surgery.

Both adhesions and hernias usually obstruct the small bowel. Large bowel obstruction is most commonly due to cancer, usually of the colon but sometimes of the ovary. Volvulus of the bowel, either large or small, initially causes obstruction but strangulation rapidly supervenes if the bowel is not untwisted.

If there is doubt about the reality or the severity of an obstruction contrast radiograph examination is always helpful. A small bowel enema is a useful way to decide whether small bowel obstruction needs surgical treatment. A conventional barium enema will help to plan an operation for large bowel obstruction and will identify patients with pseudo-obstruction.

Rare causes of obstruction such as gallstone ileus, bolus obstruction from worms or foodstuffs, an obturator hernia, and malrotation are usually diagnosed at laparotomy, although there is enormous satisfaction in making such a diagnosis in advance. This is sometimes possible if the diagnosis is considered and the relevant symptoms and signs are sought.

Acute retention of urine also presents with abdominal distension and pain but the patient will volunteer that he or she has not passed urine for some time. Most patients are male and elderly. The tense distended bladder is easy to see and to feel and there is instant relief when a catheter is passed. The soft flaccid bladder associated with chronic retention, which sometimes presents acutely, is much more difficult to feel or to percuss, and ultrasound is often of help. Sympathomimetic drugs, sometimes in small doses in proprietary medicines, and constipation are two factors which can precipitate acute retention in susceptible people.

Patients with pelvic peritonitis are sometimes sent into hospital with a diagnosis of acute retention. The inflammation has led to dehydration and severe oliguria along with lower abdominal pain and distension. The true diagnosis is often not made until a catheter is passed and only a small amount of urine is retrieved.

Abnormal contraction of smooth muscle causes the regular and intermittent pain which is called colic. Within the abdomen only the gut, the renal tract, the uterus, and the biliary tract cause such a pain. The site and the distribution of the pain are different in each case and so they are easy to tell apart except, sometimes, for right-sided renal colic and biliary pain. The most severe intestinal colic accompanies gastroenteritis, although it is a classic sign of small bowel obstruction. Stones are the common cause of renal and biliary colic, although blood clot and pus can cause ureteric colic quite as severe as that due to a stone. Most, but not all, women know when they are pregnant and are about to deliver a baby, but uterine colic due to a miscarriage or even severe dysmenorrhoea sometimes presents as acute abdominal pain.

A ruptured abdominal aortic aneurysm, acute haemorrhagic pancreatitis, and acutely ischaemic bowel are the three common conditions causing a patient to present at hospital severely shocked immediately after the onset of acute abdominal pain. Shock is also seen in any patient with significant intra-abdominal pathology who presents some time after the onset of their symptoms, but the history will usually identify the delay and lead to the correct diagnosis.

A ruptured aneurysm is never easy to feel in a hypotensive patient and severe acute pancreatitis is sometimes only diagnosed at laparotomy. Patients with dead bowel inside the abdomen look and sometimes smell as though they are dying, which in one sense is correct. Those with strangulated gut usually show signs of obstruction, while a mesenteric embolus is easy to diagnose if the patient is fibrillating but very difficult if the thrombus lies on the endocardium after a silent myocardial infarct.

Occasional patients with abdominal pain who present acutely have a mass in the abdomen on palpation. The first rule is to try and define the organ from which the lump arises and then to describe all the physical characteristics of the lump, remembering particularly to test for pulsation. An ovarian cyst and an obstructed loop of bowel are common causes of a lump; apart from these, lumps are rare. They can arise from any intra-abdominal organ. Most obscure abdominal masses are subjected to intense preoperative investigation; in reality they all require exploration at laparotomy, when their nature and their origin will be revealed. Even then all the preoperative predictions are sometimes found to be wrong.

The injured patient with a rapidly distending abdomen and signs of exsanguination obviously requires an immediate laparotomy. However, it is often difficult to decide whether a trauma victim has significant intra-abdominal injury. Most such patients have other injuries, a few are unconscious, and alcohol is a factor in some. Complaints of abdominal pain should always be taken seriously, even though few patients can give a good history and abdominal examination is limited because the normal responses are impaired. Some abdominal injuries, particularly intestinal injuries, do not become apparent immediately: repeated reassessment is therefore essential.

It is always worth establishing the mechanism of injury as this can give some indication of the likely damage. Direct blunt trauma in a car crash or from the handlebars of a pedal cycle, kicks, punches, and falls from a height can all rupture the kidney, the spleen, or the liver. A lap and diagonal seat belt will tear the small bowel mesentery if the lap belt lies too tight across the abdomen. A fractured pelvis can perforate the bladder, and is often associated with rupture of the urethra.

The feasibility of a complete examination will vary but the abdominal surgeon should certainly pay attention to the general condition of the patient, the other injuries sustained and, in particular, to the chest. Unexplained and persistent tachycardia and hypotension despite active resuscitation implies continuing haemorrhage. If the site of bleeding is not visible the blood is either in the chest or in the abdomen. Fractured ribs can damage abdominal organs as well as the lungs.

Peritoneal lavage is a useful indicator of the need for a laparotomy. Frank blood, intestinal contents, or heavily blood-stained lavage fluid clearly indicates the need for a laparotomy. The kidney is the organ most frequently injured by blunt abdominal trauma, and intravenous urography is essential in the management of traumatic haematuria. Too many surgeons have removed an injured but solitary kidney. Computed tomography of the abdomen can identify retroperitoneal injuries which are not apparent on clinical examination and will also demonstrate the extent of an injury within a solid organ.

In civilian life penetrating abdominal wounds are less common than blunt injury, but most of the former will require surgical exploration. It is usually impossible to tell how far a wound has penetrated and no simple investigation, except perhaps laparoscopy, is helpful.

Common sense must always be applied: in the trauma victim it is almost always better to ‘look and see’ rather than to ‘wait and see’ if there is a continuing possibility of a significant intra-abdominal injury.

Nine of 10 children with abdominal pain either have acute appendicitis or non-specific abdominal pain (Table 4) 402. The other child is likely to have either a urinary tract infection or intussusception. Urinary infection presents with the classical symptoms of fever, frequency, and stinging pain on micturition. The diagnosis is only confirmed when pus cells are found on microscopy of a correctly collected urine sample and a significant number of pathogenic bacteria grow on culture. In contrast, intussusception rarely presents with all the classical signs of colicky abdominal pain, a palpable abdominal mass, and redcurrant jelly stools. Any infant—most patients are under 5 years of age—who develops severe intermittent acute abdominal pain manifested as acute episodes of screaming should be suspected of harbouring an intussusception. An ultrasound examination will confirm the diagnosis and an enema either with barium, which can also make the diagnosis, or with air may also reduce the intussusception.

Two conditions which may cause non-specific abdominal pain in children deserve particular comment. Constipation is often said to be the most common cause of abdominal pain, although few such patients reach hospital. In young children, in whom a rectal examination is to be avoided, it is perhaps permissible to make the diagnosis on a plain abdominal radiograph, although even this is unnecessary if faecal masses can be felt in the abdomen. Secondly, examination of the ears and the throat are as essential as palpation of the abdomen in young children: both tonsillitis and otitis media can present with abdominal pain.

Acute appendicitis and non-specific abdominal pain are the most common causes of abdominal pain all over the world. Disease patterns vary, however, and few patients with non-specific pain are admitted to hospital in the third world.

Of the specifically tropical diseases only two are of real concern. Firstly infestation with worms is a significant cause of abdominal colic, and sometimes of intestinal obstruction. Secondly, pain and tenderness in the right upper quadrant are more likely to be caused by amoebic hepatitis with a liver abscess rather than acute cholecystitis. Amoebic colitis can present with all the signs of peritonitis but the typical shallow shaggy ulcers seen on sigmoidoscopy should make the correct diagnosis clear, When the inflammation is confined to the caecum (‘typhlitis’) the signs mimic acute appendicitis.

This rare condition can mimic intra-abdominal pathology. The haematoma develops from rupture of the inferior epigastric artery in the lower half of the abdomen. Pregnant women are particularly affected; there is sometimes a history of injury, and the right side is affected twice as often as the left.

The onset of the pain is acute and it is often accompanied by nausea and vomiting. There is marked tenderness in the iliac fossa, and it is easy to misdiagnose appendicitis. Two physical signs will reveal the true diagnosis. It may be possible to show that the tenderness, and the swelling if there is one, is confined to the abdominal wall. Secondly bruising of the skin may be visible. Sometimes this only appears a few days later and it is often at a distance from the site of maximum tenderness.

Once diagnosed the haematoma will slowly resolve with rest; in patients who undergo surgery the diagnosis becomes apparent as the abdominal incision is made. It is then worth tying off the bleeding vessel; this is also needed in the rare patient in whom the haemorrhage does not stop.

Similar bleeding sometimes arises from spontaneous rupture of an intercostal artery. The dramatic bruising of the abdominal wall spreading round from the lower thorax, often in a segmental distribution, is unmistakable. The haematoma will normally resolve with rest.

Small epigastric, periumbilical, and tiny incisional hernias which usually contain only extraperitoneal fat can be very hard to find and are surprisingly painful, even without strangulation of the contents. Epigastric hernias often cause peculiar digestive symptoms, which can lead to diagnostic confusion.

Many patients who need an operation for acute abdominal pain also have other medical problems, and some strictly medical conditions are treated by surgeons simply because they usually present as a surgical problem. Here we consider those patients who present with an acute abdomen but the underlying cause requires the care of a physician.

Inflammation of the pleura at the base of the lungs can present as abdominal pain because of irritation of the lower intercostal nerves which supply the abdominal wall. Pneumonia in the very young and the very old, and pulmonary embolism at any age, can present in this way. Usually there are some symptoms and signs referrable to the chest, although in infants pneumonia may only be diagnosed on a chest radiograph.

Occasionally the chest abnormality is itself secondary to pathology below the diaphragm. Rarely, perforated diverticular disease produces minimal symptoms from the peritonitis but leads to some subphrenic infection. Pneumonia supervenes because of poor diaphragmatic movement and by the time the patient comes to the hospital the abdominal symptoms are long since forgotten. Conservative management of the abdominal problem is usually appropriate anyway, so the diagnostic error is rarely of serious consequence. Cholecystitis can sometimes present in the same way.

Occasionally pain from an acutely distended liver is the presenting feature of right-sided heart failure, although other physical signs are usually also present. Myocardial infarction is said, on occasion, to present with continuous epigastric pain. The main distinguishing feature is the complete lack of any epigastric tenderness.

Warfarin and digoxin are the two common drugs that cause abdominal pain, although neither is as widely used as in the past. Digoxin toxicity presents with abdominal pain and vomiting. Warfarin anticoagulation can cause a spontaneous retroperitoneal haemorrhage with an associated paralytic ileus and severe abdominal pain radiating to the back. Haematuria is a clue to the diagnosis and the prothrombin time is usually excessively prolonged.

Very rarely diabetes mellitus first manifests itself with acute abdominal pain. The diagnosis is rarely difficult because the patient, who is usually an adolescent, is obviously severely ill with impairment of consciousness and marked dehydration. Sometimes the sweet ketotic smell typical of hyperglycaemic ketoacidosis pervades the whole examination room.

Inflammation of the liver can sometimes cause intrahepatic cholestasis. If the acutely distended liver is also painful it is easy to think that the patient has extrahepatic obstructive jaundice. Both drugs and viral infections can produce this clinical picture, although the matter is resolved as soon as the bile ducts are of normal size on an ultrasound examination.

About half the children with Henoch-Schönlein purpura complain of abdominal pain. The pain is due to bleeding into the wall of the bowel and this can lead to an intussusception. The diagnosis will only be missed if the characteristic skin rash is absent. Spontaneous and painful intra-abdominal haemorrhage is also a feature of haemophilia.

There are a few strange people in every society who enjoy being admitted to a hospital. Some of them are looking for a bed, some want drugs, some like attention, and a few need help. Most are well known to their own local medical services and some even achieve a wider notoriety.

Such patients usually present themselves in the accident and emergency department and give a dramatic history of some acute abdominal event. Renal colic, a potential rupture of the spleen, and a fall astride a bar are all popular. They can always simulate the necessary physical signs. They often appear in very severe pain. A tiny cut on the lip, a finger, or the genitalia is quite sufficient to produce haematuria. They will also claim to be allergic to intravenous contrast media so that it is impossible to confirm a diagnosis of renal colic. Many are admitted for observation and it may be some days before the true diagnosis is disclosed.

Many such patients are simply too good to be true. They fail to realize that it is exceptionally rare for any patient to have every recorded symptom of their chosen diagnosis. Most patients give an address which is a long way away and some describe themselves as lorry drivers. Other clues are early and repeated demands for opiate analgesia and many previous admissions to hospital. In the classical case of Munchausen's syndrome there are also multiple abdominal scars.

Even when the diagnosis is suspected it is difficult to confront the patient since doctors are naturally loath to accuse patients of lying. It is better to investigate the background by telephoning doctors who have treated the patient in the past, either in a hospital or in the community, and making it clear to the patient that this is taking place. Demands for pain relief should be met by the offer of non-opiate analgesics. Most patients simply leave the hospital, often without telling the staff, once this process of enquiry starts or when they fail to obtain the service which they were seeking.

The patient with acute abdominal pain always presents the surgeon with the challenge of making the correct diagnosis. Without that diagnosis the right treatment cannot be offered.

Making the correct diagnosis is never easy. It demands attention to detail in taking the history and examining the patient and clarity of thought in analysing the information that is obtained. Investigations may help but in many places in the world there are no facilities for further investigation. There the management of every patient depends entirely on the clinical skills of the doctor. Finally every surgeon should remember that for every five patients admitted to a hospital with abdominal pain at least one patient has his or her diagnosis changed during the course of the admission and two leave the hospital without being given a cause for their pain. This does not mean that there was no cause. It does mean that our skill in making a diagnosis needs to be improved.

Adams ID, et al. Computer aided diagnosis of acute abdominal pain: a multicentre study. Br Med J 1986: 293: 800–4.

Bailey H, Love M. Vermiform appendix. In: A Short Practice of Surgery. 12th edn. London: Lewis, 1962: 978–1002.

Campbell JPM, Gunn AA. Plain abdominal radiographs and acute abdominal pain. Br J Surg 1988; 75: 554–6.

Cope Z. The Early Diagnosis of the Acute Abdomen. 9th edn. London: Oxford University Press, 1948.

de Dombal FT. Diagnosis of Acute Abdominal Pain. 2nd edn. Edinburgh: Churchill Livingstone, 1991.

De Dombal FT, Leaper DJ, Horrocks JC, Staniland JR, McCann Ap. Human and computer-aided diagnosis of abdominal pain: further report with emphasis on performance of clinicians. Br Med J 1974: 1: 376–80.

Editorial. Analgesia and the acute abdomen. Br Med J 1979; 279: 1093.

Field S, Guy PJ, Upsdell SM, Scourfield AE. The erect abdominal radiograph in the acute abdomen: should its routine use be abandoned? Br Med J 1985; 290: 1934–6.

Gallegos N, Hobsley M. Abdominal pain: parietal or visceral? J R Soc Med 1992; 85: 379.

Gray DWR, Collin J. Non-specific abdominal pain as a cause of acute admission to hospital. Br J Surg 1987: 74: 239–42.

Gray DWR, Dixon JM, Collin J. The closed eyes sign: an aid to diagnosing non-specific abdominal pain. Br Med J 1988; 297:837.

Hockerstedt K, Airo I, Karaharju E, Sundin A. Abdominal trauma and laparotomy in 158 patients. Acta Chir Scand 1982: 148: 9–14.

Hoffman J, Rasmussen OO. Aids in the diagnosis of acute appendicitis. Br J Surg 1989; 76: 774–9.

Jones DJ. Appendicitis. Br Med J 1992; 305: 44–7.

Jones PF. Active observation in management of acute abdominal pain in childhood. Br Med J 1976; 274: 551–3.

Jones PF. Emergency Abdominal Surgery. 2nd edn. Oxford: Blackwell Scientific Publications, 1987.

McAdam WAF, Brock BM, Armitage T, Davenport P, Chan M, de Dombal FT. Twelve years' experience of computer-aided diagnosis in a district general hospital. Ann R Coll Surg 1990; 72: 140–6.

Macartney FJ. Diagnostic logic. Br NMed J 1987; 295: 1325–31.

Masters K, Levine BA, Gaskill HV, Sirinek Kr. Diagnosing appendicitis during pregnancy. Am J Surg 1984; 148: 768–71.

Mohammed R, Goy JA, Walpole BG, Brentnall EW, Miller T. Munchausen's syndrome. A study of the casualty ‘Black books’ of Melbourne. Med J Aust 1985; 143: 561–3.

O'Shea B, McGennis A, Cahill M, Falvey J. Munchausen's syndrome. Br J Hosp Med 1984; 31: 269–74.

Paterson-Brown S, Vipond MN, Simms K, Gatzen C, Thompson JN, Dudley HAF. Clinical decision making and laparoscopy versus computer prediction in the management of the acute abdomen. Br J Surg 1989; 76: 1011–13.

Paterson-Brown S, Eckersley JRT, Sim AJW, Dudley HAF. Laparoscopy as an adjunct to decision making in the ‘acute abdomen’. Br J Surg 1986: 73: 1022–4.

Paterson-Brown S, Vipond MN. Modern aids to clinical decision making in the acute abdomen. Br J Surg 1990; 77: 13–18.

Puylaert JBCM, et al. A prospective study of ultrasonography in the diagnosis of appendicitis. N Engl J Med 1987; 317: 666–9.Raheja SK, McDonald P, Taylor I. Non-specific abdominal pain—an expensive mystery. J R Soc Med 1990; 83: 10–11.

Sawyers JL, Williams LF, eds. The acute abdomen. Surg Clin N Am 1988; 68: 233–476.

Sherck JP, McCort JJ, Oakes DD. Computed tomography in thoraco-abdominal trauma. J Trauma 1984; 24: 1015–21.

Siddiqui MN, Abid Q, Qaseem T, Hameed S, Ahmed M. ‘Spontaneous’ rectus sheath ahematoma: a rare cause of abdominal pain. J R Soc Med 1992; 85: 420–1.

Stewart J, Finan PJ, Courtney DF, Brennan TG. Does a water soluble contrast enema assist in the management of acute large bowel obstruction: a prospective study of 117 cases. Br J Surg 1984; 71: 799–801.

Thomson HJ, Jones PF. Active observation in acute abdominal pain. Am J Surg 1986; 152: 522–5.

Williamson RCN, Cooper MJ, eds. Clinical Surgery International. Vol. 17. Emergency Abdominal Surgery. Edinburgh: Churchill Livingstone, 1990.

Winsey HS, Jones PF. Acute abdominal pain in children: analysis of a years admissions. Br Med J 1967: 1: 653–5.

Wyatt GM, Spitz HB. Ultrasound in the diagnosis of rectus sheath haematoma. JAMA 1979; 241: 1499–1500.