![[ Visit Client Website ]](images/banner.jpg)
BACKGROUND: Regional Health Information Organizations (RHIOs) in New York City (NYC) collect electronic data from healthcare facilities and facilitate patient lookup via a web portal. Access to RHIOs for case investigation of reportable diseases could improve the efficiency and timeliness of investigation by reducing the need for on-site medical record review and/or provider interviews. We evaluated the RHIO patient lookup interface as a source for case investigation data and compared it with traditional medical record review.
METHODS: We compared RHIOs A and B with traditional case investigation for previously reported and investigated cases of legionellosis, a disease for which chart review is routinely conducted. Cases included NYC residents who were reported from hospitals participating in either RHIO. All cases from RHIO A hospitals between 1/01/2013 and 12/31/2013 were included. Of cases from RHIO B hospitals between 7/01/2013 and 06/31/2014, a subset was randomly selected. After determining whether the individual patient record was available in the RHIOs, we searched for key data points routinely collected through traditional chart review. We calculated completeness of data elements from both sources and compared values for concordance; data were classified as matching if values from both sources were populated and equivalent. Dates were considered equivalent if they differed by <= 1 day.
RESULTS: Results were similar for the two RHIOs. Of 56 cases from RHIO A, 53 (95%) were located in the RHIO and evaluated. Of 131 cases from RHIO B, 37 cases were selected; 36 (97%) were found in the RHIO. Laboratory results, chest imaging results, and hospitalization data were highly concordant (matching over 90% of the time) and complete; 89% of patients found in RHIOs had urine antigen tests and 93% had chest imaging results available. Past medical history and symptoms were least likely to be found; among individual symptoms, fever was most likely to be captured (completeness was 51%, compared with 87% from traditional review). Onset date was also poorly captured. Overall, completeness was higher from traditional chart review than from the RHIOs.
CONCLUSIONS: In NYC, two major RHIOs currently provide highly complete and accurate hospitalization, laboratory, and radiologic information. RHIO patient lookup portals would be more useful if narrative summaries including symptom and onset data were available. NYC RHIOs have the potential to greatly improve the efficiency of public health case investigation; further evaluation is needed as they evolve.