The Alarming Rate of Major Disruptive Events in Primary Care Practices in Oklahoma

The purpose of this study was to document and assess the rates of major disruptive events in a cohort of primary care practices in Oklahoma. The authors argue that disruptive events can adversely affect continuity of care and quality improvement efforts in primary care.

In the introduction, the authors state that quality improvement directors and practice facilitators routinely report the negative effects of major disruptive events on efforts for improvement and change. However, the discussion group pointed out that the authors do not provide evidence from prior studies that disruptive events do in fact impact primary care practices. Although it is understandable that a major disruptive event certainly could disrupt the workflow and care provided by an office, this still seems like an assumption that the authors make. For example, sometimes change in clinicians or key staff might lead to a positive outcome if the prior staff worked inefficiently or had poor communication.

The study utilized practices that were part of the EvidenceNOW trial. This trial provided interventions such as peer consultation, on site information technology assistance and monthly performance feedback in order to assess the effectiveness of external quality improvement support in improving patient health. During this intervention, an electronic questionnaire was developed for the enrolled practices that elicited information on major disruptive events. Practice facilitators chose from a provided list of disruptive events or described events with free text/writing. The study excluded a total of 58 practices of the original 208, 19 of which dropped out which is a fairly large number. The discussion group pointed out that the authors do not provide more details regarding why so many practices dropped out and whether the 58 total excluded practices were different in some way that could have affected the results of the study. In addition, the data was collected by self-report which raises the question of whether there is response bias and to what degree. Practices undergoing major disruptive events may be less likely to report adverse outcomes.

Furthermore, the practice population that was assessed in this study involved a significant number of Native American tribes. There were 2 Indian Health Services practices and 10 tribal practices total. The discussion group pointed out that this might affect the external validity of the study because health centers serving these communities may have a harder time recruiting physicians and experience a higher rate of major disruptive events. In addition, the discussion group preferred to see a more detailed breakdown of what types of disruptive events were reported by small, medium and large practices. The group felt that there could be more discussion of how the classification of "disruptive events" was determined. Different practices might have their unique interpretation of disruptive events and each disruptive event might affect primary care practices differently. For example, if a primary care practice lost an older clinician but a younger practitioner more familiar with the EHR system was hired, this would actually be a positive change in improving effective patient care. The implementation of an EHR, while at first can be burdensome, typically results in better patient care when the office fully commits to it. Providers are able to quickly pull up patient records as well as lab and imaging results rather than waiting for the results to be sent over via other methods such as fax.

The results of the study reveal that 63% of small primary care practices, 70% of medium-size and 27% of large practices reported at least 1 major disruptive event. The authors discussed that the reason medium sized practices reported greater staff changes might be due to medium sized practices lacking the proper resources that larger practices have and the cohesiveness of smaller practices. Among the discussion group, there were differing opinions on the validity of these results. Some members of the group agreed that it makes sense the large practices were better able to withstand the major disruptive events. If there are more clinicians and resources, one could reason that there is more support to withstand these changes. However, because practice facilitators were only present for a few hours every two weeks to collect data, there could be significant underreporting especially in larger practices that have more personnel and thus more opportunities for change. It is also interesting that the authors point toward cohesiveness as a possible reason why the small practices had fewer disruptive events compared to the medium sized ones. Although small practices reported fewer disruptive events, the discussion group mentioned that these practices could have been more affected by the changes due to having fewer resources and professional support.

The group discussed a few recommendations to improve the study if it were to be conducted again. This study relied solely on self-report questionnaires from practice facilitators and staff. It would be interesting to compare the reports of these individuals versus clinicians and see if their responses differ. Clinicians are directly involved in patient care and thus might be able to better describe the effects of major disruptive events especially with regards to continuity of care. In addition, the discussion group would have liked to see a more detailed comparison of the impact of disruptive events on quality improvement goals between the different sized practices rather than just a few anecdotes. Perhaps a scaled rating system can be included in the questionnaire that assesses which disruptive events affect primary care practices more.

Overall, the discussion group agreed that the study would have been more meaningful if it also included more detailed analysis of the impact of disruptive events, root causes of these events, and methods to prevent and reduce their negative impact on health systems. Rather than focusing solely on identifying the presence of disruptive events, it is equally important to assess how healthcare systems are affected by these events. Future studies might be able to assess this aspect further and the results can be utilized to guide primary care practices to respond to change that maximizes quality patient care. In addition, this study was performed solely in Oklahoma and it is difficult to know whether the results are applicable to the general population. More research is required to assess whether similar disruptive events can be identified in other locations.

Competing interests: None declared