Diagnosis related groups: the effect of input error.

Abstract

American studies on Diagnosis Related Groups (DRG) coding errors have suggested large error rates. However, we are aware of no Australian data on the extent and significance of errors in completion and encoding of front sheet diagnosis/procedure information on DRG grouping of hospital separations. With the imminence of casemix funding in Australia, this topic takes on a greater significance. We therefore undertook a pilot research study to assess this. At each of three hospitals in the Hunter Area Health Service, NSW (one teaching, one acute general district hospital and one rural general practitioner hospital), we attempted to sample 150 records. A committee reviewed the charts and assessed coding, DRG and Major Diagnostic Category (MDC) error rates. These were 8.1-9.6% for DRG allocation and 3.4-5.2% in MDC allocation. Using current casemix funding data, one of the hospitals would have suffered a funding shortfall for inpatient services of nearly $2 million or 2.6% of total budget at that time. If these data, extracted after an exercise to reduce errors in DRG and front sheet coding, are representative, then there are major implications for hospital inpatient funding if casemix funding is introduced. Major educational exercises are called for the further in-depth research may prove useful.