A 12-year analysis of Malmquist total factor productivity in dialysis facilities

This study examined total factor productivity of dialysis facilities in Greece over a 12-year period, using nationally representative panel data. Data Envelopment Analysis (DEA) was used to compute Malmquist productivity indices, which were decomposed into technical efficiency change and technological change. The sample consisted of 73 dialysis facilities operating throughout the entire study period (1993-2004), corresponding to 97.3% and 58.9% of all facilities in the first and last study years respectively. Production variables were nursing staff and dialysis machines (inputs) and dialysis sessions (output). The DEA model was input-oriented allowing for constant returns to scale (CRS). Technical efficiency change was decomposed into scale efficiency change and variable returns to scale (VRS) "pure" technical efficiency change. Mean overall efficiency, throughout the study years, ranged from 39.6 to 63.1% with an all-time average of 56.7%, and only 2-4% of the facilities were fully efficient in each study year. Productivity indices indicated year-by-year progress or regress up to 5%, but the efficiency and technological components differed, in some cases, by as much as 30%. Although interesting subperiod effects were observed, conclusions could not be generalized for the entire study period due to alternating trends. We suggest that preliminary insight to productivity in this sector has been obtained, but particular subperiods must be isolated and further investigated.