Health IT Study: Good Try, But No Cigar

A highly publicized study of Texas hospitals found that the use of health IT significantly improves quality and efficiency. But some observers question whether the results reflect the use of electronic records or simply the correlation between health IT adoption and a culture of excellence in some hospitals.

Published in the Archives of Internal Medicine, the study assessed the level of automated clinical decision support in 41 urban hospitals. The researchers found that for every 10-point increase in the automation of notes and records, there was a 15 percent decline in the mortality rate for all conditions studied. Those conditions included myocardial infarction, heart failure, coronary artery bypass graft, and pneumonia. The study also found that greater automation lowered complications and hospital costs.

This sounds like really great news, especially in light of the Obama Administration’s bold efforts to promote health IT. But there are dissenters. Dr. Deborah Hammond of Ridgewood, NJ, for instance, wrote a letter to Modern Healthcare noting that better-financed institutions that do high volumes of certain procedures may have better outcomes, regardless of their degree of automation. “EHRs are important for fully integrated, quality care over time, but we should be careful to not over-anticipate the results of what will be a very expensive transition to electronic records,” she wrote.

Kurt Williamsen’s letter to Modern Healthcare was more pointed. “The article implies that computerizing records will reduce mortality and costs, but while the article does show some correlation, it does not show causation. It just may be that mortality goes down because hospitals that can afford to computerize their records can also afford to pay to have the best doctors and have the newest operating-room equipment.”

The study’s authors say that they did adjust the data for for each hospital’s size, profit margin, and ownership status, which they believe could affect its quality orientation. But they admit that “the [study] design does not consider a number of organizational confounders that could explain superior clinical outcomes, most notably a hospital’s emphasis on safety and quality.”

This study took an important step toward understanding how physicians’ use of information technology improves quality and safety. But it did not establish that automated decision support alone is sufficient to raise quality and reduce costs. In fact, a well-known experiment with computerized physician order entry, by Cedars-Sinai Medical Center, flopped because physicians did not want to use the system. The success of health IT in individual hospitals–as well as physician practices–depends on a number of complex factors that are endemic to the local healthcare environment.

A good example is the experience of the Geisinger Health System, which has had an electronic health record since 1995. This large integrated system has introduced a number of innovations to improve its performance, as detailed in a recent Health Affairs paper. It is piloting a “medical home” project that includes $1800 monthly care coordination payments and additional stipends for physicians who care for Medicare patients. The early results from this pilot show a 20 percent reduction in admissions and a 7 percent drop in total medical costs. Geisinger also has a comprehensive disease management program. And it has pioneered a risk-based approach to reengineering acute care in its hospitals, in which providers agree to follow a specific set of best practices and refund half of their charges for CABG surgeries if their readmission rates exceed the group average.

All of these innovations, the authors of the Health Affairs article point out, depend on the use of EHRs. And yet, they note, “Only within the past few years has Geisinger begun to leverage key benefits from its electronic infrastructure—after a long period of implementation, adoption and usability comfort was created among users. Much of today’s policy discussions imply that EHRs will rapidly transform care delivery. The Geisinger experience suggests that this is not the case but, rather, that EHR adoption is the beginning of a long care-transformation journey.”