Read Core Topics in General & Emergency Surgery: Companion to Specialist Surgical Practice Online
Authors: Simon Paterson-Brown MBBS MPhil MS FRCS
Core Topics in General & Emergency Surgery: Companion to Specialist Surgical Practice (23 page)
Attempts to improve preoperative diagnosis and early management of patients with acute abdominal pain are continuingly being made; this section discusses all the current available techniques and the evidence for their incorporation into emergency surgical practice. For the purposes of most studies looking at acute abdominal pain, the broad definition is taken as ‘abdominal pain of less than 1 week's duration requiring admission to hospital, which has not been previously treated or investigated’. However, this must be accepted as a fairly loose definition.
For an emergency team system to work efficiently the surgical team must have rapid access to diagnostic blood tests and appropriate imaging, which should include plain and contrast radiology, both diagnostic and interventional (percutaneous drainage and biopsy), ultrasound (US) and computed tomography (CT). Furthermore, plain radiography evaluated by senior radiologists substantially enhances senior surgical assessment of patients with acute abdominal pain, resulting in reduced surgical admissions.
30
All these modalities are discussed below.
Many studies have looked at the spectrum of patients admitted to hospital with acute abdominal pain and the approximate percentage represented by each condition is now well understood. Figures from one study
31
appear to be fairly representative (
Box 5.3
). In this study the 30-day mortality in 1190 emergency admissions was 4%, with a perioperative mortality of 8%. Not surprisingly, the mortality rate was age related, with perioperative mortality in patients below 60 years being 2%, rising to 12% in those 60–69 years and reaching 20% in patients over the age of 80 years. Laparotomy for irresectable disease was the most common cause of perioperative mortality (28%), with ruptured abdominal aortic aneurysm (23%), perforated peptic ulcer (16%) and colonic resections (14%) all being associated with significant perioperative mortality.
Box 5.3
Conditions that may present with acute abdominal pain
Non-specific abdominal pain (NSAP) (35%)
Acute appendicitis (17%)
Intestinal obstruction (15%)
Urological causes (6%)
Gallstone disease (5%)
Colonic diverticular disease (4%)
Abdominal trauma (3%)
Abdominal malignancy (3%)
Perforated peptic ulcer (3%)
Pancreatitis (2%)
Conditions contributing 1% or less
Exacerbation of peptic ulcer
Ruptured abdominal aortic aneurysm
Gynaecological causes (these may go unnoticed as NSAP)
Inflammatory bowel disease
Medical conditions
Mesenteric ischaemia
Gastroenteritis
Miscellaneous
Reproduced from Irvin TT. Abdominal pain: a surgical audit of 1190 emergency admissions. Br J Surg 1989; 76:1121–5. © British Journal of Surgery Society Ltd. Reproduced with permission. Permission is granted by John Wiley & Sons Ltd on behalf of the BJSS Ltd.
What stands out from all the studies on acute abdominal pain published over the last few decades is the high incidence of non-specific abdominal pain (NSAP), with published figures of 40% or more.
32
NSAP usually reflects a failure of diagnosis, as many of these patients do have a cause for the pain and it has been shown that further investigations, such as laparoscopy, can reduce the overall incidence of NSAP to around 27%.
33
Some authors have examined this diagnosis of NSAP further and describe a certain number of alternative conditions that could be related (
Box 5.4
),
32
including abdominal wall pain
34
and rectus nerve entrapment.
35
In some cases of NSAP, detection of abdominal wall tenderness (increased abdominal pain on tensing the abdominal wall muscles) may be a useful diagnostic test.
36
Possible causes of abdominal wall pain are also given in
Box 5.4
. The major problem with making a diagnosis of NSAP is in missing serious underlying disease, and the late Tim de Dombal demonstrated that 10% of patients over the age of 50 years who were admitted to hospital with acute abdominal pain were subsequently found to have intra-abdominal malignancy.
37
Half of these patients had colonic carcinoma and the major concern was that 50% of the patients who were subsequently proved to have intra-abdominal cancer were discharged from hospital with a diagnosis of NSAP.
Box 5.4
Causes of non-specific abdominal pain
32
Viral infections
Bacterial gastroenteritis
Worm infestation
Irritable bowel syndrome
Gynaecological conditions
Psychosomatic pain
Abdominal wall pain
34
Iatrogenic peripheral nerve injuries
Hernias
Myofascial pain syndromes
Rib tip syndrome
Nerve root pain
Rectus sheath haematoma
Acute gynaecological conditions such as pelvic inflammatory disease and ovarian cyst accidents are another group of diagnoses that may often be included under the umbrella of NSAP, simply because of failure to take a good history or examination or even perform a thorough pelvic examination, whether digitally, ultrasonographically or at operation. In one study from a general surgical unit, gynaecological causes represented 13% of all diagnoses in a consecutive series of all emergency admissions (both male and female) initially presumed to be ‘surgical’ in origin.
38
As many of these patients present with ‘query appendicitis’, accurate assessment is essential if unnecessary operations are to be avoided, and even then the diagnosis may remain hidden unless the surgeon examines the pelvic organs once a normal appendix has been found. However, with the increased use of diagnostic laparoscopy, discussed later in this chapter, these conditions are now being recognised by the emergency surgeon with much greater frequency. Early recognition and appropriate treatment of pelvic inflammatory disease may help to avoid potentially serious long-term sequelae and must be encouraged.
39
Indeed, the condition of Curtis–Fitz-Hugh syndrome, when transperitoneal spread of pelvic inflammatory disease produces right upper quadrant pain due to perihepatic adhesions, is now well recognised and care must be taken to differentiate this from acute biliary conditions.
40
Although much is made of possible ‘medical’ causes of acute abdominal pain in surgical textbooks, the incidence of conditions such as myocardial infarction, lobar pneumonia and some metabolic disorders is extremely small, though many still masquerade as NSAP. However, the possibility of such conditions must still be borne in mind during the clinical assessment of all patients with acute abdominal pain: one study has shown that 19 of 1168 children (1.6%) admitted to hospital with acute abdominal pain had pneumonia as the sole cause of symptoms.
41
It is therefore still extremely important to recognise these medical conditions when they do present, before exploratory surgery, as the mortality can be significantly increased.
42
In the early 1970s, de Dombal et al.
43
in Leeds and Gunn
44
in Edinburgh developed a computer program based on Bayesian reasoning that produced a list of probable diagnoses for individual patients with acute abdominal pain. They demonstrated that the accuracy of clinical diagnosis could be improved by around 20%, and a subsequent multicentre study confirmed this finding.
45
Furthermore, these studies showed that there was a reduction in the unnecessary laparotomy rate and bad management errors (patients whose surgery is incorrectly delayed). When the reasons for the improvement in diagnostic accuracy associated with the use of computer-aided diagnosis (CAD) were examined, there appeared to be three main factors involved: (i) peer review and audit, which is invariably associated with improved results in most aspects of medical management;
46
(ii) an educational factor related to feedback;
47
and (iii) probably of greatest significance, the use of structured data sheets on to which the history and examination findings were documented before being entered into the computer program. One study went on to demonstrate that the diagnostic accuracy of junior doctors improved by nearly 20% when they used structured data sheets alone, without going on to use the CAD program.
48
The same study also demonstrated that medical students assessing patients with the structured data sheets and then using the CAD program reached similar levels of diagnostic accuracy to qualified doctors. Other studies have since confirmed similar improvements in clinical decision-making following the introduction of these data sheets.
49
The message from all these studies is clear: a good history and examination remain essential for both diagnostic accuracy and good clinical decision-making, and the use of a structured data sheet helps the clinician to achieve this objective.
8,
9,
45
The aim of both the history and examination is to determine a diagnosis and clinical decision. There are undoubtedly specific features associated with all acute abdominal conditions that are well established; however, it remains the ability to identify the presence or absence of peritoneal inflammation that probably has the greatest influence on the final surgical decision. In other words, the presence or absence of guarding and rebound tenderness, and a history of pain on coughing, correlates well with the presence of peritonitis.
50
The differential diagnosis of acute appendicitis from NSAP is always difficult, particularly in children, and both guarding and rebound tenderness are significantly more likely to be present in acute appendicitis.
51
There was always great emphasis in the past on the importance of a rectal examination in patients with suspected acute appendicitis to elicit tenderness within the pelvis. However, when rebound tenderness is detected in the lower abdomen, as evident by pain on gentle percussion, further examination by rectal examination has been shown to provide no new information.
52
Rectal examination can therefore be avoided in such patients and reserved for those patients without rebound tenderness or where specific pelvic disease needs to be excluded. Measurement of temperature has also been shown to be relatively non-discriminatory in the early assessment of the acute abdomen.
53
Urgent urinary microscopy should be carried out on anyone with any symptoms that could relate to the urinary tract and it is a good principle to dip test the urine on admission of every patient with acute abdominal pain.
After the initial assessment (history and examination) of patients with acute abdominal pain, steps should be taken towards resuscitation, pain relief and further diagnostic tests as required. There is very good evidence to support the early administration of opiate analgesia in patients with acute abdominal pain, and a recent Cochrane systematic review confirmed that patient comfort was improved without detrimentally affecting surgical decision-making.
54
The early administration of opiate analgesia in patients with acute abdominal pain improves patient comfort without adversely affecting clinical decision-making.
54
Once the initial assessment has been completed, the surgeon will reach a differential diagnosis and, perhaps more importantly, a clinical decision: early operation definitely required, early operation definitely not required or need for early operation uncertain. Clearly, further investigations in the first category are unlikely to influence management, with the exception of a serum amylase level, which may reveal acute pancreatitis.
55
Further investigations in the group in which the surgeon considers early operation is not required can be organised on a more leisurely basis, and it is not surprising that it is in the group in which the surgeon is uncertain as to whether early operation is required that most difficulty exists.
56
Most of the uncertainty relates to ‘query appendicitis’, particularly in the young female, but also involves patients with intestinal obstruction and the elderly patient, in whom the diagnosis of mesenteric ischaemia must always be considered.
57
Early diagnosis in patients with mesenteric ischaemia is particularly important as survival after surgery is much better in those with venous thrombosis than those with arterial thrombosis.
58
In the assessment of the role of subsequent investigations in the acute abdomen, it is important to identify their potential influence on clinical decision-making rather than evaluating them purely on diagnostic potential.
Although blood tests are often useful as a baseline, their influence on the diagnosis of acute abdominal pain remains unclear, with the exception of serum amylase
55
and increasingly serum lipase
59
or acute pancreatitis (see also
Chapter 8
). Studies examining the influence of white cell concentration
60
and C-reactive protein
61,
62
in patients with ‘query appendicitis’ have concluded that serial white cell counts are useful (compared with a single measurement). Although isolated C-reactive protein levels may also be fairly non-discriminatory, when they are interpreted with white cell count and both are normal, acute appendicitis is unlikely.
63
Overall, inflammatory markers are poor discriminators of conditions such as appendicitis when looked at individually, but when combined and used with history and clinical findings of peritoneal irritation they achieve a high discriminatory power.
64
Thus, routine measurement of the white cell count and C-reactive protein in patients with acute abdominal pain can be justified, not only for a baseline with which to compare subsequent levels depending on clinical progress, but also to be interpreted along with all other clinical and biochemical findings.
Liver function tests are of course an essential investigation in the early assessment of the acute abdomen where acute biliary disease is suspected.
65
The other area that has attracted great interest in the role of blood tests for aiding diagnosis in the acute abdomen is intestinal ischaemia, whether from strangulated obstruction or mesenteric ischaemia and infarction. Estimation of acid–base status to assess the degree of metabolic acidosis is often a late change and measurement of serum phosphate, lactate, kinase, creatine, lactate dehydrogenase, alkaline phosphatase, diamineoxidase and porcine ileal peptide have all been shown to be unreliable.
58
A recent study, however, has demonstrated the use of a combined clinico-radiological score for predicting the risk of strangulation in small-bowel obstruction.
66
This will be discussed in more detail later.
