Technological progress as a means to improve hospital productivity - easy to observe but difficult to measure

Presenter: Kirsi Vitikainen, National Institute for Health and Welfare (THL)

Abstract

Background: DRGs (Diagnoses Related Groups) have originally been designed for hospital reimbursement purposes, but have also been used as a casemix adjustment tool in hospital output and productivity measurement. From the perspective of productivity measurement the DRG cost weighting for outputs might be inappropriate as costs do not reflect the real value of treatments. During the last decade hospital service production has increasingly been shifted from expensive inpatient settings to less expensive outpatient settings (day surgical and ambulatory care settings), which has made the total DRG weighted output appear lower over time. This is because cost based weighting gives less value to outpatient procedures compared to inpatient procedures even if the outcome of both procedures is the same. An alternative to cost weighting would be to attach a common, value based weight to procedures which give the same clinical outcome irrespective of treatment settings.

Objectives: The aim of this study is to explore the difference that cost and value based weightings cause to time series output and productivity measurement. We use two patient groups which have increasingly been treated in outpatient settings instead of inpatient settings in the period 2001–2006 as examples.

Data and methods: The Finnish hospital benchmarking database is used to form two datasets from years 2001–2006 with patient cases with the same diagnosis or procedure that can be treated in either inpatient or outpatient settings. Inguinal hernia patients represent a surgical patient group and thrombophlebitis patients a medical patient group. The former group is formed based on a procedure code and consists of 67,992 patient cases, and the latter group based on a diagnosis (ICD-10 codes I801–I803) and consisting of 19,920 patient cases.

Output quantity index and productivity index approach following Laspeyres is used to explore the difference between cost and value based weightings.

Results: Output and productivity indexes are highly sensitive to the choice between cost and value based weightings for both inguinal hernia and thrombophlebitis. For both diseases the difference in total weighted output and productivity between the two weighting techniques grows over time, mirroring the increased share of outpatient treatment. For inguinal hernia, a one per cent increase in the share of outpatient treatment accounts for approximately 0.7 per cent greater difference in the total weighted output between cost and value based weightings. The equivalent figure with respect to productivity is 0.5 per cent. For thrombophlebitis, a one per cent increase in the share of outpatient treatment accounts for approximately 1.3 per cent greater difference in the total weighted output and 2.2 per cent greater difference in productivity between the weighting techniques.

Conclusions: The use of cost weights provides a poor indication of output and productivity growth as it gives less value to outpatient treatment compared to inpatient treatment even if the outcome of both treatments are often the same. In national accounts value based weighting should be used instead of cost based weighting.