Early assessment of the acute abdomen

Conditions associated with abdominal pain

The list of potential causes of the acute abdomen is extensive. An exhaustive account is not the purpose of this chapter, and readers will find the latest guidance on diagnosis and management of common conditions elsewhere in this book or the Companion to Specialist Practice series. It is now over 30 years since Irvin et al. documented the diagnoses presenting with acute abdominal pain to a UK general surgery department. At that time, the most common diagnosis (over a third of the cohort) was non-specific abdominal pain (NSAP), twice as frequent as the next most common condition (acute appendicitis, 17%).

NSAP may be defined as ‘pain for which no immediate cause can be found following examination and baseline investigations and specifically does not require surgical intervention’. A variety of causes have been proposed ( Box 11.1 ). Historically, clinicians worried that a label of NSAP might mask serious undiagnosed pathology. However, data from children with abdominal pain suggests NSAP is a safe diagnosis: a large retrospective study of > 3000 admissions with NSAP in children over 20 years found a ‘missed’ appendicitis rate of only 0.2%. A record linkage study of a cohort of > 250 000 children with NSAP 1999–2011 from English national data found that only 5.8% were subsequently hospitalised for bowel disorders, the most likely conditions being appendicitis, IBD and IBS. In adults, increased use of diagnostic imaging in the past 10–20 years (see below) suggests that a diagnosis of NSAP is likely to be much more secure than when de Dombal et al. observed that 10% of patients over 50 years old labelled as NSAP presented subsequently with an intra-abdominal malignancy (most commonly colorectal cancer).

Irvin’s audit has been updated in a 2009 multicentre study of 1021 adult patients presenting with acute abdominal pain to EDs in the Netherlands ( Table 11.1 ). Although the frequency of individual diagnoses remained broadly similar over time, it is notable that the frequency of NSAP was substantially less in the Dutch study and no doubt reflects the more advanced imaging strategies employed. However, that depends on referral pathways: the proportion of non-surgical causes of acute abdominal pain admitted to a New Zealand surgical unit increased during the same 10-year period despite greater use of diagnostic imaging.

A small number of medical conditions can present with acute abdominal pain, and although uncommon are mentioned here for the benefit of surgeons in training. Inferior myocardial infarction, lower lobar pneumonia and some metabolic disorders can all be excluded by examination and/or basic investigations (ECG, chest radiograph and serum glucose); failure to recognise them is associated with significantly increased morbidity and mortality.

Initial assessment: history, examination and simple tests

A careful medical history and clinical examination remains the keystone of initial assessment and should lead to formulation of a differential diagnosis from the conditions listed in Table 11.1 . In a review of abdominal pain assessment errors in the ED, failure of history taking was one of the biggest contributors. Age is an important determinant of likely diagnoses; the differential of, for example, right iliac fossa pain in teenagers is quite different in octogenarians.

It is now well established that adequate (usually opiate) analgesia for patients presenting with an acute abdomen does not mask abdominal signs and is not detrimental to surgical assessment or decision-making.

However, as previous generations of general surgeons knew well, in managing the acute abdomen there is an important distinction between assessing urgency and making an accurate diagnosis . Although some conditions are recognised reliably by clinical assessment (particularly acute diverticulitis and small bowel obstruction ), in general the accuracy of clinical diagnosis in the acute abdomen is only moderate. The Acute Abdominal Pain Study Group found that diagnostic accuracy was less than 50% with substantial interobserver variation, particularly in eliciting physical signs, but distinguishing urgent from non-urgent conditions was more reliable. In the previous era, the main decision for the general surgeon was when to operate immediately, when to observe and when not to operate at all. A precise diagnosis was less of a priority, partly because the diagnostic armamentarium was limited. Consequently, the prevailing negative laparotomy rate at the time was considerable (‘better to look and see than wait and see’). In the patient with peritonitis and septic shock, it may still be argued that a precise diagnosis is less important than rapid intervention to resuscitate and achieve source control (and the diagnosis) by laparotomy. However, for patients in whom the need for operation is less obvious, in modern practice a precise preoperative diagnosis has important implications:

• Subspecialisation in general surgery is now the norm in many countries, hence diagnosis is important for onward referral to the appropriate subspecialty (which may be in a different hospital). The emergence of the emergency general surgeon in UK practice has embedded this process in many hospitals.

• To avoid unnecessary admission to hospital in an increasingly resource-limited service. Although time is a key determinant in the evaluation of the acute abdomen, and active observation is a well-established and safe practice, hospital bed occupancy is costly.

• For selection of the appropriate treatment option depending on severity (operative vs non-operative; laparoscopic vs open surgery).

A number of studies have sought to improve the accuracy of clinical diagnosis by systematic documentation of clinical variables to develop risk prediction scoring systems, applied most frequently to acute appendicitis (e.g. Alvarado score, Appendicitis Inflammatory Response score ). A comprehensive comparison of appendicitis risk prediction models in a large UK ‘snap-shot’ audit of > 5000 patients has recently been published (overall, the Adult Appendicitis Score performed best, with a failure rate of 3.7%). In current practice, these scoring systems probably have greatest application in allowing less experienced or non-surgical clinicians to triage patients that may safely avoid hospital admission, to select patients for diagnostic imaging and to provide a framework for clinical reassessment. ^,

Initial investigations