What diagnoses may make patients more seriously ill than they first appear? Mortality according to the Simple Clinical Score Risk Class at the time of admission compared to the observed mortality of different ICD9 codes identified on death or discharge.

BACKGROUND The Simple Clinical Score (SCS) determined at the time of admission places acutely ill general medical patients into one of five risk classes associated with an increasing risk of death within 30 days. The cohort of acute medical patient that the SCS was derived from had, on average, four combinations of 74 groupings of ICD9 codes. This paper reports the ICD9 codes associated with the different SCS risk classes and identifies those ICD9 codes with a greater observed mortality than that of other patients in the same SCS risk class. DESIGN Observational study. SETTING A small Irish rural hospital. METHODS The 30-day mortality rates of the 74 commonest ICD9 groupings coded at the time of discharge of 9214 consecutive acutely ill medical patients were compared with the mortality rates associated with their SCS risk class determined at the time of their admission. RESULTS There was no difference between the observed and the predicted mortality rates for very low risk patients regardless of ICD9 groupings, even though several of these patients suffered from all but two of the 34 ICD9 code groupings associated with an increased risk of death. Within the remaining four risk classes only 14 ICD9 groupings had an observed mortality greater from that of all other patients in the same SCS risk class. CONCLUSION The Simple Clinical Score (SCS) determined at the time of admission identifies patients at very low risk of death regardless of what diagnoses are subsequently made during their hospitalisation. Nevertheless, patients with a very low risk of death according to their SCS risk class may still have a life-threatening condition that requires treatment in hospital. For higher risk patients only 14 ICD9 code groupings were associated with an observed mortality greater than that of others in the same SCS risk group.

[1]  M. Charlson,et al.  Clinical assessments as predictors of one year survival after hospitalization: implications for prognostic stratification. , 1988, Journal of clinical epidemiology.

[2]  J. Kellett,et al.  The Simple Clinical Score predicts mortality for 30 days after admission to an acute medical unit. , 2006, QJM : monthly journal of the Association of Physicians.

[3]  J. Kellett Acute hospital medicine in Ireland — past, present and future , 1999 .

[4]  L. Lind,et al.  Comparison of the rapid emergency medicine score and APACHE II in nonsurgical emergency department patients. , 2003, Academic emergency medicine : official journal of the Society for Academic Emergency Medicine.

[5]  K. Godfrey Comparing the Means of Several Groups , 1985, Medical Uses of Statistics.

[6]  J. Kellett,et al.  Differences between self-referred and physician-referred hospital admissions , 2005, Irish journal of medical science.

[7]  J. Kellett,et al.  The diagnoses and co-morbidity encountered in the hospital practice of acute internal medicine. , 2007, European journal of internal medicine.

[8]  The appropriateness of admissions and the influences on a decision to admit. , 2000, Journal of quality in clinical practice.

[9]  Giles Morgan,et al.  Confidential inquiry into quality of care before admission to intensive care , 1998, BMJ.

[10]  C. Subbe,et al.  Validation of a modified Early Warning Score in medical admissions. , 2001, QJM : monthly journal of the Association of Physicians.

[11]  A. W. Kemp,et al.  Medical Uses of Statistics. , 1994 .