original article S T Ward1 S Gilbert M Mulchandani W V Garrett Department of Surgery & Anaesthetics, Medway Maritime Hospital, Windmill Road, Gillingham, Kent ME7 5NY Correspondence to: Mr Stephen Thomas Ward Fryern Wood, Chaldon, Caterham, Surrey, CR3 5AR Tel: + 44(0)7904 216421 Fax: +44(0)1268 583888 email: drsteveward@yahoo com UNNECESSARY INVESTIGATIONS IN PATIENTS WITH ACUTE PANCREATITIS: ARTERIAL BLOOD SAMPLING AND SERUM BIOCHEMISTRY Background: Most patients admitted with acute pancreatitis undergo arterial blood gas sampling (ABG) to calculate the modied Glasgow score (MGS) and serum amylase and liver function tests (LFTs) are requested frequently during admission This study aims to assess the necessity of these investigations Materials & methods: A retrospective study of all patients attending a district general hospital with a diagnosis of acute pancreatitis (AP) between November 2005 and November 2006 was performed Patients were identied from clinical coding data, excluding those with serum amylase levels of 55 years White cell count >15×109/l Blood glucose >10mmol/l Urea >16mmol/l Arterial oxygen partial pressure =3 factors detected within 48 hours Amylase requests following diagnosis are therefore inappropriate and waste resources Serial LFTs may not be necessary on a daily basis but there is little evidence to justify how frequently they should be performed Total number of necessary amylase requests 65 Total number of unnecessary amylase requests 162 On admission 63 To formulate a MGS 44 MATERIALS AND METHODS Following ERCP We performed a retrospective study of all patients admitted between November 2005 and November 2006 with a clinical diagnosis of AP Patients were identified using clinical coding data (OPCS 4.4/ICD-10 code: K85) The diagnosis of AP was confirmed by interrogation of the biochemistry reports database, identifying all reports of a serum amylase level greater than 550iU/l (>x5 upper limit normal for our laboratory) over the same time period Only patients with diagnostic amylase levels of >550iU/l were included in the study to increase the likelihood of correct diagnosis The case notes of the identified patients were retrieved to obtain demographic data, confirm the diagnosis and acquire ABG reports and the corresponding pulse oximeter derived oxygen saturations at the time of the ABG sampling The full complement of blood results for each patient on each day of admission was retrieved from the computer results database system The hypothesis that MGS scores with the PaO2 component excluded can be used to assess severity of disease was tested The degree of metabolic acid-base disturbance for each arterial blood sample was stratified as none (base excess >-2), minimal (-2>base excess> -4), mild (-4>base excess> -6), moderate (-6>base excess> -9), marked (-9>base excess> -13) or severe (base excess550iU/l on admission by cross-referencing the biochemistry results database A total of 63 patients (39 female, 24 male) were included in the study, mean age 59 years (range 16-91) Case notes were retrievable in all but two cases © 2008 Surgeon 6; 5: 282-7 LFT Not requested for =72 51 hours Total number of necessary LFT requests 161 Total number of unnecessary LFT requests 168 Two hundred and twenty-seven amylase results, 329 LFT results, 95 ABG reports and 80 MGSs were retrieved and analysed An MGS was calculated on admission for 56 (89%) patients The distribution of all MGSs can be seen in Figure The severity of pancreatitis as indicated by the maximum MGS within the first 48 hours of admission was mild in 37 (61%) cases, severe in 17 (28%) cases and not recorded in the remaining (11%) cases The underlying aetiology was gallstones in 44 (70%) cases, alcohol in (10%) cases and unknown in 13 (20%) cases Only eight ABGs were found to have PaO2 values less than kPa, a result that could be predicted in all cases by pulse oximeter-derived oxygen saturations of under 95% The MGS excluding the PaO2 parameter was always two or more in cases where the PaO2 was less than kPa (see Figure 2) The MGS excluding the PaO2 parameter was always three or more in cases where acid-base disturbance exhibited more than a mild metabolic acidosis (see Figure 3) Sixty-three of the 80 MGSs were scored during the first 48 hours of admission Analysing only these ‘early’ scores still shows that the MGS excluding the PaO2 parameter was always two or more in cases where the PaO2 was less than kPa and always three or more in cases where acid-base disturbance exhibited more than a mild metabolic acidosis A total of 162 unnecessary amylase requests and 168 unnecessary LFT requests were made according to our defined criteria This equates to 2.6 unnecessary amylase requests and 2.7 unnecessary LFT requests the royal colleges of surgeons of edinburgh and ireland | 283 Fig Distribution of all MGS scores excluding the PaO2 parameter per admission The costs of biochemical assays in our laboratory are outlined in Table It can be seen that the total cost of unnecessary biochemistry investigations across the study period was £83.40 or £1.32 per admission DISCUSSION The MGS is used to stratify severity of AP, a score of three or more indicating severe disease The advantages of the MGS are that the parameters are few in number and can therefore be committed to memory and easily applied The score is useful before the initial 48 hour period if the criteria are fulfilled Ranson’s score is composed of 11 criteria, some evaluated on day one and others on day two The score is not valid until 48 hours have elapsed and the criteria thresholds depend on the pancreatitis aetiology.13 The Acute Physiology and Chronic Health Evaluation II (APACHE II) score can be calculated on a day-to-day basis but is complex and lacks ease of application outside an intensive care setting, leading investigators to formulate simplified scores.14 The serum C-reactive protein (CRP) correlates well with severity of disease but its usefulness is again delayed by 48 hours, sensitivity rising from 284 | Table Cost of biochemistry requests Biochemistry sample handling charge Urea & electrolytes (U&Es) Amylase Lactate dehydrogenase (LDH) £3.00 £0.10 £0.10 £0.10 Liver function tests Alanine transaminase (Alt) Alkaline phosphatase (Alp) Total bilirubin Albumin (Alb) C-reactive protein (CRP) £0.4 £0.10 £0.10 £0.10 £0.10 £0.60 the royal colleges of surgeons of edinburgh and ireland © 2008 Surgeon 6; 5: 282-7 Fig Percentage of ABG reports with PaO2