Californians wanting to select a hospital or surgeon on the basis of publicly available information on quality will find serious shortcomings in the available data.

This was the finding from a review of 18 organisations in California, USA reporting 333 measures of healthcare quality. Shortcomings were at several levels, although foremost was that all the organisations examined quality at the hospital level—none at the level of the individual surgeon, group of surgeons or health plan.

For 21 procedures, accounting for 21% of the surgical procedures, structure, process and outcome measures were looked for by the researchers. Organisations reported structure measures (such as annual hospital volume) for 12 procedures. None reported any process measures. Outcome measures were reported for 19 procedures, the most commonly reported being death in hospital and the major complication rate.

Six of the 10 most common non-obstetric procedures, including hysterectomy and cholecystectomy, had no reported quality measures at all. Furthermore because of the time lag between collecting and reporting data, most current measures reflected care that had been delivered between two and five years previously.

The situation is set to improve by 2005 when three new publicly available quality measures are expected relating to hip fracture, carotid endarterectomy and coronary artery bypass grafting. For the last of these the quality measure will include risk adjusted deaths specific to individual surgeons.

An accompanying commentary commented that there are essentially three problems to the task of improving health care by publishing outcomes from healthcare providers. The first is to find outcomes that provide good comparable information, allowing for differences in case mix and with sufficient power that differences between providers do not arise by chance.

The second is to make sure that this information is used to genuinely improve the quality of care provided by those underperforming—and not just a change in their reporting. Finally, as illustrated by the review from California, the third problem is to find measures that comprehensively capture information about health care which is meaningful to the individual and, importantly, provide this information in a timely way.

.