Less is More: HIE Prevents Repeat Tests

Health Information Exchange (HIE) lets doctors see all of their patients’ records online, no matter where patients choose to go for care. If a patient has a doctor they really like in one hospital, but wants their labs drawn and prescriptions filled at a more convenient location in their neighborhood, then winds up in the ER across town after falling at a shopping mall, , and their bone specialist can see the results of the X-rays done at the ER when the patient goes for follow-up. Without HIE, the doctor’s notes, medication list, labs, and X-ray results can be scattered across town in several different computer systems.

Source: Flickr/CC

Source: Flickr/CC

This study determined if HIE decreased repeat imaging tests (X-rays, ultrasounds, and CT scans) in an 11-county region in New York. They concluded that every 36 times a provider checks a patient’s records in the HIE, 1 less repeat image is done.

For several reasons, this study probably underestimates how much an HIE can reduce repeat images: 1) On average, doctors in this study checked their patients’ HIE record after 12% of the images were done. More repeat tests could have been prevented if doctors used the HIE more. 2) The study includes all health care visits, but most of them were at clinics. The clinic visits may dilute the importance of the HIE to hospitals or ERs which are more likely to repeat tests to get the quick result they need to make a treatment decision on the spot. 3) To decide whether the HIE was responsible for preventing a repeat image, the study looked at whether any doctor had opened the patient’s HIE record between when the first image was taken and when the repeat image was taken. However, if one doctor looked at the result in the HIE a week after the test and didn’t repeat it because they could see the result, but another doctor didn’t look at the HIE two weeks later and repeated the test, this was counted as a failure of the HIE to prevent the repeat image instead of the success it actually was when the first doctor used the HIE.

There are some good reasons to repeat images which this study could not capture. The study called any test that was repeated within 3 months a duplicate, but after 1-4 weeks, doctors may repeat a test to see whether the patient’s problem is getting better or worse. Even when obtaining repeat tests, doctors with access to HIE can compare new tests with older ones, improving patient care and further illustrating the benefit of HIE systems.

commentary by Laura Medford-Davis, MD

Abstract

Objectives Medical imaging, which is expensive, is frequently repeated for the same patient within a relatively short period of time due to lack of access to previous images. Health information exchange (HIE) may reduce repeat imaging by facilitating provider access to prior images and reports. We sought to determine the effect of an HIE system on the occurrence of repeat imaging. Study Design and Methods We conducted a cohort study of adult patients who consented to participate in a community-based HIE system in an 11-county region in New York. Using data from 2009-2010, we linked log files of provider HIE usage to administrative claims data from 2 commercial health plans. Using generalized estimation equations, we measured the association between HIE system access and repeat imaging within 90 days. Results Of 196,314 patients in the cohort, 34,604 (18%) of patients had at least 1 imaging procedure, which was equivalent to a rate of 28.7 imaging procedures per 1000 patients. Overall, 7.7% of images were repeated within 90 days. If the HIE system was accessed within the 90 days following an initial imaging procedure, imaging was significantly less likely to be repeated (5% repeated with HIE access vs 8% repeated without HIE access, p<.001). HIE system access reduced the adjusted odds of a repeat image by 25% (95% CI, 13%-35%). Conclusions Use of the HIE system to access previous patient information was associated with a reduction in repeated imaging. PMID: 25811815

Vest, JR. et al. AJMC. 2014; 20: eSP16-24.