Patients classification System

tools for
Cost management

tools for
Quality management

tools for

4 Why to choose US

4 implementation


4 contact

4 Logic

4 diagnoses

4 procedures

4 inpatients

4 outpatients

4 insurance

4 readmissions

4 reoperations

4 complications

4 deaths

4 length of stay

4 day surgery

4 unjustified stays


As for any quality indicator all undesirable events cannot be avoided even with optimal care. Hospital medicine is always associated to risks, which has to be analyzed with statistical tools. Thus, the quality is questionable if observed complications rate is higher than expected rates, computed in taking into account the case mix of eligible discharges

Complication excess (dCP) is given by the difference between observed (CP1) and expected complication rates (CP0): dCP = CP1-CP0.


An excess of complication might reflect a greater propensity to code them all, including mild complications such as redness of the skin. In such cases, we should not observe a lengthening of the stay. A length of stay excess is observed when lengths of stay exceed expected ones more among complicated cases than uncomplicated ones.

with complications without complications

Observed length of stay LS1C LS1

Expected length of stay LS0C LS0


Length of stay excess dLS = (LS1C LS0C) - (LS1 LS0).

All excess are expressed in positive values; if negative, they are equalized to zero.


Observed rate (CP1) Expected rate (average (CP0), minimal (CP0min), maximal (CP0max)

Rate ratio (RCP): observed rate (CP1) / expected rate (CP0)
Swiss rate (CP0CH): observed rate in Switzerland
Adjusted rate (CP1a): rate ratio Swiss rate (CP0CH)


A (good) rate ratio < 0.8 and significantly lower than expected rate (CP1 < CP0min) and no length of stay excess (dLS= 0)
B (normal) 0.8
rate ratio 1.2 or not significantly different from CP1 (CP0min < CP1 < CP0max) or no length of stay excess (dLS= 0)
C (to be analysed) rate ratio > 1.2 and significantly higher than expected rate (CP1 > CP0max) and length of stay excess (dLS> 0)

SQLape s..r.l. and Yves Eggli, 2014. Last update: 20.05.2014