Link Between Volume and Health Care Outcomes Studied
The Institute of Medicine has published a summary of a workshop on interpreting the volumeoutcome relationship. Workshop discussions were based on two background papers: a cross-disciplinary review of published research and a synthesis of the literature relevant to health care policy.
The literature review examined 88 studies involving eight procedures or conditions and found that higher volume was related to better outcomes in three fourths of the studies. However, some low-volume providers had excellent outcomes and some high-volume providers poor outcomes. The difference between low- and high-volume providers narrowed as procedures became well established, suggesting that volume as an indicator of quality may be most useful in relation to new technology. Few of the studies detailed the structures or processes underlying the apparent relationship.
Presenters of the second background paper cast the volumeoutcome relationship in the broader context of quality improvement. Volume-related strategies for improving care could be regulatory, competition based (e.g., report cards for consumers, selective referrals by purchasers), or aimed at upgrading the skills of health care professionals. Potential problems with selective referrals based on volume were discussed, including decrements in quality of care at high volumes, patients' preference for receiving care close to home, and price increases at high-volume hospitals as they gain market power.
One participant noted that policies directing patients to high-volume centers would affect many people, because a substantial proportion of patient care takes place in lower-volume facilities. Another stated that process measures would be preferable to volume as indicators of quality. Further research on the volumeoutcome relationship, said another, would give individual care providers information for improving quality. The need for clinical data, not just administrative data (e.g., hospital discharge data), was stressed.
While agreeing that volume may be the best available indicator of quality for some conditions, the participants thought efforts were needed to explain the effect of volume and to look at other quality indicators in conjunction with volume. A wide-ranging list of research topics was identified. Some participants thought research confirming the volumeoutcome relationship should be made available to the public for use in health care decision-making but stressed that the context for interpreting this information should be made clear.