Is the Colombian health system reform improving the performance of public hospitals in Bogotá?

Many countries are experimenting with public hospital reform - both increasing the managerial autonomy with which hospitals conduct their affairs, and separating 'purchaser' and 'provider' sides of the health system, thus increasing the degree of market pressure brought to bear on hospitals. Evidence suggesting that such reform will improve hospital performance is weak. From a theoretical perspective, it is not clear why public hospitals should be expected to behave like firms and seek to maximize profits as this model requires. Empirically, there is very slight evidence that such reforms may improve efficiency, and reason to be concerned about their equity implications. In Colombia, an ambitious reform programme includes among its measures the attempt to universalize a segmented health system, the creation of a purchaser-provider split and the transformation of public hospitals into 'autonomous state entities'. By design, the Colombian reform programme avoids the forces that produce equity losses in other developing countries. This paper reports the results of a study that has tried to track hospital performance in other dimensions in the post-reform period in Bogotá. Trends in hospital inputs, production and productivity, quality and patient satisfaction are presented, and qualitative data based on interviews with hospital workers are analyzed. The evidence we have been able to collect is capable of providing only a partial response to the study question. There is some evidence of increased activity and productivity and sustained quality despite declining staffing levels. Qualitative data suggest that hospital workers have noticed considerable changes, which include greater responsiveness to patients but also a heavier administrative burden. It is difficult to attribute specific causality to all of the changes measured and this reflects the inherent difficulty of judging the effects of large-scale reform programmes as well as weaknesses and gaps in the data available.