iPLEDGE Weaknesses: Is It Time to Address the Flaws?

include: 1) Two prescriptions for isotretinoin were given during a single visit but only one prescription was documented; 2) Between office visits, a prescription was called in over the phone without documentation; 3) The pharmacy divides isotretinoin from a single prescription into two dates in order to satisfy insurance criteria; 4) The patient obtains an additional prescription by means other than above. Our personal experience is that certain insurance policies allow either a maximum number of isotretinoin tablets dispensed at one time (usually 90), or a certain number of days of medication (no more than 21 days per prescription); however, this alone should not result in additions, since any further dispensing of drug is ideally corroborated by an additional prescription in the medical chart.

We have termed the remainder of discrepancies, with restarts and additions discounted, as "true misses," and account for 13.4% of all charts. One explanation relies on patient compliance and accuracy of reporting (eg, patient reports that s/he is taking medication as prescribed but does not actually fill the prescription, this would result in a "miss"). Though patient non-compliance likely accounts for a portion of the discrepancy, there is no way to verify the data, and it seems unlikely that non-compliance alone accounts for all of the "true misses." Another possible explanation includes prescriptions obtained

from online/out-of-country pharmacies. In this situation, a patient would be taking the medication as prescribed reflecting the corresponding chart documentation while there would be no similar record in iPLEDGE. It is important to note that while this situation is possible, in our study would be an unlikely scenario as patients required a minimum of one month medication dispensed through iPLEDGE to be included in our study. Additionally, if patients were receiving medication from a non-regulated source, one would expect documentation in the chart. In no instance was the latter observed. Other factors (eg, pharmacy dispensed medication without corresponding entry in iPLEDGE database) are thought to contribute as well.

This study compared data from three sites; two of the sites (Einstein and Downstate) used paper medical charts, while Austin used electronic medical records (EMR). Although all three sites were substantially discordant with iPLEDGE, Austin had fewer discrepancies than either of the two sites that utilized paper charts. Though we did not set out to study the difference between accuracy of electronic versus paper charts, this sizable difference in percentage of observed discrepancies (17.2% vs. 34.0%) is difficult to ignore.

EMR has potential benefits in areas of clinical, organizational, and societal outcomes.¹⁰ In one study designed to evaluate completeness and uniformity of health record data entry for history and physical examination, electronic records were found to be superior to paper charts.¹¹ Another large multicenter study of 461 test results from 200 charts found that results managed with an EMR were more often in the right place in the chart, and had more clinician signatures, interpretations, and patient notifications documented.¹²

Our study reinforces the findings of improved accuracy for electronic charting, as the Austin site (ie, with EMR) had notably less discordance (16.8% lower rate), compared to the sites utilizing paper. A few noteworthy differences in electronic charting included the relative ease of data retrieval, as prescriptions were consistently located within a specific area of the chart, and records of telephone notes were captured in Austin charts but not in the paper charts. Since telephone notes frequently explained when a patient did not fill their prescription for isotretinoin, classification as a "miss" was avoided in the absence of a corresponding dispensed dose in iPLEDGE. Neither site utilizing paper charts had a reliable way to track additional correspondence outside of the standard office visit. Additionally, the Austin EMR was notable for having a cumulative dose tracker, which could help account for the lower number of additions at this site.

From its inception, iPLEDGE was not specifically designed to function as an electronic database. Nonetheless, to "prevent