IS THERE A RELATIONSHIP BETWEEN RAISED PLASMA OSMOLARITY AND THE ONSET OF ACUTE PANCREATITIS

A.K. Banerjee , R.Jain , J.Coore , T.W. Balfour Depts. of Surgery , Univs. of Nottingham and Sheffield , U.K. INTRODUCTION The aim of this study was to study the relationship of plasma osmolarity with the onset of pancreatitis , and also whether there was any relationship of osmolarity with a predicted severe attack or aetiology' . The study design comprised a retrospective series of 111 consecutive patients with acute pancreatitis admitted to the City Hospital Nottingham between 1987 and 1992 . The diagnosis was based on the presence of a serum amylase level > 300 units/litre ( upper limit of normal 90 units/litre ) with a compatible clinical picture . The aetiology of the 111 cases of pancreatitis was predominantly gallstones and alcohol in a similar ratio to that reported in other series' . The following laboratory results were obtained : sodium , potassium , creatinine , serum amylase , alanine aminotransferase,lactate dehydrogenase, glucose urea,calcium ,serum albumin levels, haematocrit , white cell count , arterial pH and Po2 . It was therefore possible to place each patient into a predicted mild or severe attack group using modified Ranson criteria' in addition to calculating plasma osmolarity using the formula 1.89 x Na i 1.38 x K i 1.03 x Urea i 1.08 x Gluc i 7.45. Calculated osmolarity was compared to those obtained from a control group . The latter consisted of 60 non-diabetic patients with similar age and sex distribution presenting with abdominal pain and serum amylase levels within the reference range . The control patients were consecutive emergency admissions encompassing some serious intraabdominal conditions necessitating major surgery and who were therefore considered as unwell as some of the group with severe pancreatitis .=ist ics Since the data is not normally distributed , Spearman Rank correlation for non-parametric data was used to investigate the correlation between osmolarity and amylase . The Mann Whitney U test was used to compare the osmolarities between groups . Coefficients of variation were calculated using regression analysis from Statgraphics programme ( STSC Inc., 2115 East Jefferson St. , Rockville , Maryland 20852 , USA ).of the 60 patients in the control group , 35 ( 58% ) were male . The age ranqe was 22 to 92 years ( median 52 ) , and the mean osmolarity was 293 ( interquartile range 286 297 ) mmol/l. There was no difference between the controls with mild nonspecific abdominal pain or those with severe conditions such as intestinal perforation . In the mild pancreatitis group , 58 out of 90 were male (64%) , with an age range of 19 to 91 years ( median 56 ) . The mean plasma osmolarity of this group was 290 ( interquartile range 281 299 ) mmol/l . There was no relationship between osmolarity and aetiology of pancreatitis . 21 patients had a predicted severe attack of pancreatitis of which 7 were fatal . Thirteen of this group were male ( 62% ) I and the age range of the group was 42 to 87 years ( median hi ) . The mean plasma osmolarity of this qroup was 313 (interquartile range 293 333) mmol/l . Thls was a strongly significant findinq ( Statistics using Mann-Whitney U test Mild v. Controls z= 1.421 , p=O.1 : Mild v . Severe z = -3.54, p < 0.001 : Severe v.