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

INTRODUCTION

In response to rising health care costs, the changing epidemiology of illnesses, and opportunities created by the advancement of information technologies, federal and state governments, insurance companies, and employers are applying pressure on primary care practices to transform the way they provide care. Common requirements involve implementation of electronic health records (EHRs), instituting population management and care management systems, and restructuring practice reimbursement from a fee-for-service model to value-based reimbursement models. These dramatic changes have created anxiety and uncertainty, which could be expected to result in some degree of organizational instability.^1–7

Quality improvement directors and practice facilitators, who function as change agents helping primary care practices manage these challenges, routinely report that major disruptive events, such as relocations and changes in ownership, clinicians, and key staff, interfere with their efforts to effect change and sustain improvements. Surprisingly, the actual rate and impact of major disruptive events have rarely been documented or published. Although they can be difficult to track and report in a great number of practices over extended periods, there may also be reluctance on the part of implementation researchers to acknowledge the fragility of primary care practices and adverse effects of that fragility on transformation efforts.

As a part of the Healthy Hearts for Oklahoma study, a 3-year quality improvement project that enrolled primary care practices, we collected information about major disruptive events. Our objective was to document the number and type of disruptive events and to generate questions about impact and causation.

METHODS

Our study tracked 6 categories of major disruptive events occurring in a set of primary care practices in Oklahoma during a 2-year period. Data were collected by surveys of practice managers and staff, who reported retrospectively on major disruptive events that occurred within 1 year before study enrollment and obtained prospectively through reports entered by practice facilitators as they observed the workflow in their assigned practices.

All aspects of the project were approved by the University of Oklahoma Health Sciences Center Institutional Review Board, as well as the institutional review boards of Absentee Shawnee Tribe, Cherokee Nation, Chickasaw Nation, Choctaw Nation, Oklahoma City Area Indian Health Service, Integris Heath, and Ponca Tribe; all participants provided informed consent.

RESULTS

Of the 208 practices, we excluded 31 practices that had been in existence for less than 1 year at time of enrollment, 8 practices that did not complete enrollment paperwork, and 19 practices that enrolled but dropped out during the first year of the project. Three practices had been open for less than 1 year and also did not complete enrollment paperwork.

The enrolled practices were, by intention, distributed over the entire state, with greater proportions per population located in rural areas (Supplemental Appendix, available at http://www.annfammed.org/content/16/Suppl_1/S52/suppl/DC1/). Self-reported characteristics of the 208 practices, including location, ownership, and practice type, size, and specialty composition, are shown in Table 1. Numbers of practices located in metropolitan vs nonmetropolitan counties were comparable. In terms of ownership, 39% were clinician owned, 25% were hospital or health system owned, and 19% were federally qualified community health centers. One-half of practices reported having 2 to 5 clinicians in their practice, and two-thirds were single-specialty practices.

On the questionnaires, 120 (58%) practices reported having experienced 175 major disruptive events (Table 2), and 42 practices reported multiple major disruptive events. Sixteen-percent of the practices had implemented a new EHR, and 17% had installed a new billing system. Overall, 28% of practices had lost at least 1 clinician. Fifteen practices reported other events. One reported changing to a professional limited liability company, another reported implementation of the International Classification of Diseases, Tenth Revision (ICD-10) coding system, and 6 reported that additional practice locations were added. One practice reported the loss of nonadministrative nursing staff, and 2 reported the addition of a clinician.

During the first year postenrollment, 89 major disruptive events occurred in 67 (32%) unique practices (Table 2), and 20 of those practices reported multiple disruptive events. Across the entire 2-year period (retrospective plus prospective), 137 (66%) of the 208 practices experienced at least 1 major disruptive event. Comparing the second year of the study period with the first, major disruptive events related to technology or human resources occurred less frequently by more than 50%, movement of practice locations remained the same, and the number of practices that were purchased by or became affiliated with a health care system increased.

Figure 1 displays the number of major disruptive events as a function of practice size. About 63% of small primary care practices, 70% of medium-size practices (2 to 10 clinicians), but only 27% of large practices (>10 clinicians) reported at least 1 major disruptive event. Medium-size practices were somewhat more likely to experience at least 1 disruptive event than solo and larger practices (χ²= 8.75; P= .01) (Figure 1). Fifty percent of small and medium-size primary care practices experienced 1 to 2 disruptive events. More medium-size practices (19%), however, reported having 3 or more major disruptive events compared with 15% of small practices. Only 1 large practice reported 3 such disruptive events. No statistically significant associations were found between practice location, ownership, or number of clinicians and occurrence of at least 1 disruptive event during the 2-year period.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Number of practices by total number of major disruptive events.

Although we specifically excluded practices that dropped out of the study during the first year, disruptive events had a major impact on the capacity of those remaining practices in implementing and continuing their quality improvement activities, which was anecdotally supported by practice facilitators’ entries in the electronic practice records. This practice facilitator’s notes documented the effects of personnel change, the most commonly occurring disruptive event, on the practice’s ability to achieve its quality improvement goals:

We had a clinic that was part of a health system. To recruit the clinic to H2O [Healthy Hearts for Oklahoma], we had to meet several times with different levels of the organization: medical director, clinic manager, clinician, etc. Ultimately, they did decide to participate. At different times during the project, the organization has lost both the medical director and the clinic manager. After each person was replaced, the new employee made us freeze the H2O project until they could get their feet under them and decide whether the clinic could continue to participate in the project. Shutting down the project multiple times during the year of intervention not only affected the momentum of the staff participants, but it affected the PDSA [plan, do, study, act] cycles such that we couldn’t tell for sure whether the changes were having the desired effect, or whether other variables were responsible.

Another practice facilitator noted the disruptions caused by technology change:

One of my clinics is switching from a family-owned practice to being bought out by a corporate company to become the local hospital clinic. The provider does not touch a computer. The nurses do all of the documentation and scribing for the provider. These nurses have been with this provider since the beginning, and these transitions of going from paper to electronic have not been their best moments. I have heard countless stories of them wanting to go back to paper because electronic takes so much more time. They are very old school in this clinic and do not like change. They have gotten used to this system and they [are] finally used to workflow and enjoy using it. Now they are being forced to change EHRs again. This has brought out many struggles and frustration among the staff. They didn’t want the change the first time, and now they are having to change again.

These entries from the cadre of practice facilitators at Healthy Hearts for Oklahoma lend further insights into the possible impact of major disruptive events on the quality improvement efforts of primary care practices.

DISCUSSION

Our finding that two-thirds of the practices in our sample experienced at least 1 major disruptive event during a 2-year period should be alarming to everyone involved in practice transformation efforts. The fact that the practices in our sample had willingly enrolled in a quality improvement initiative and participated for at least 1 year suggests the actual rate of major disruptive events in primary care practices may be even higher.

We identified only 1 comparable study, published in 1997, involving primary care practices in Minnesota.⁸ In that study, Magnan et al reported an even higher 1-year rate of ownership changes (64%), changes in practice manager or medical director (45%), personnel changes (59%), and EHR implementation (45%) than we have.

The most frequent major disruptive events involved loss of key personnel. Personnel loss, both clinical and managerial, decreases organizational resources (eg, organizational competencies, and human capital) from which organizations may draw support in times of uncertainty.⁹ Inadequate staffing inhibits the ability of the practice to participate in improvements as well as routine training activities.¹⁰ The loss of clinical staff can be particularly problematic, because clinicians often provide both leadership and cognitive support for quality improvement.¹¹

The rate of major disruptive events did not vary across ownership types or geography (metropolitan vs rural). Even though it makes mathematical sense that practices with more clinicians and staff would be at greater risk for personnel changes, medium-size practices in our study reported greater staff changes or loss than did the solo and large practices. Solo practices are less complex and possibly more cohesive, resulting in higher levels of clinician and staff allegiance, thus avoiding certain types of major disruptive events. Medium-size practices, by contrast, may lack the resources and buffers of larger practices and the cohesiveness of small practices, making them most susceptible.

Improvements in care processes are only beneficial if they can be sustained. Although sustainability requires more study, it is likely that stability of clinicians, staff, and record systems are important, because those are the main sources of institutional memory. In the face of major disruptive events, clinician and staff training programs are likely to be essential. Factors accounting for personnel loss need to be studied to identify mitigating strategies.