Unfortunately, I can't disagree with anything you wrote. It is important that they get this right for so many reasons,…
A recent article in JAMA Health Forum caught my eye with this title: “Association of Primary Care Visit Length With Potentially Inappropriate Prescribing.” The study’s goal was to answer the question, “Are primary care physicians more likely to prescribe potentially inappropriate medications during shorter visits?” in part as a proxy for seeing whether shorter visits resulted in lower-quality care.
The authors looked at visits from 4.3 million patients, noting that “those who were younger, publicly insured, Hispanic, or non-Hispanic Black had shorter primary care physician visits.” These visits were associated with higher rates of antibiotic use for inappropriate conditions, such as upper respiratory infections. They were also associated with prescriptions for both opioid analgesics and benzodiazepines in patients who presented with painful conditions. The authors concluded that shorter visits were associated with some measures of inappropriate prescribing, but not all.
It’s a difficult study to analyze, especially looking at the demographics included in the study. We know from other research that there can be racially and ethnically associated differences in care quality. We know that black women have higher maternal mortality and less prenatal care compared to white women, and there are plenty of other examples of racial disparities in care. It’s also difficult to understand from the write-up exactly what kind of EHR data was used in the study, which was conducted from March 2022 through January 2023.
The researchers pulled a dataset from multiple states across the US that encompassed both claims and EHR data from users of the Athenahealth platform. EHR timestamp data was used, including flags for check-in, patient intake, the clinician encounter, checkout, and signoff. In some clinics, these stamps can be a poor proxy for patient visit duration, especially when there’s a lot of waiting involved or when physicians don’t appropriately change the status of visits as they move through their schedules. I would be interested to see data on the concordance of those timestamps with actual visit durations as observed in the practice before using them as a proxy.
The visit time was variable between physicians, and although the median visit length was 18.9 minutes, the range was 14.1 minutes to 24.6 minutes. There was some data I didn’t expect when looking at visit length alone. Those visits that were scheduled for 30 minutes rather than 10 received more physician attention, as one would expect. However, the difference in time spent was only four minutes for the longer appointments. That might indicate that triage algorithms or human schedulers aren’t doing a great job predicting the correct appointment slot for a given patient.
Not surprisingly, visits that had five or more diagnoses were 9.1 minutes longer than those with only one recorded diagnosis. New patient visits were 4.1 minutes longer than those with established patients. The data supported previously proven conclusions, such as female patients having longer visits than male patients and older patients having longer visits than younger patients. It also showed that patients with commercial insurance had slightly longer visits than those with Medicaid or other payers.
The researchers found a correlation between longer visits and a decreased likelihood of inappropriate antibiotic use. On the flip side, longer visits had a positive association with potentially inappropriate prescribing among adult adults, which was an interesting finding. The authors note that “many of the prescriptions that we observed may have been refills; thus, it may have taken the physician less time to refill the medication than to engage in a discussion about de-prescribing.”
The authors end by stating that there are opportunities for additional research and operational interventions for visit scheduling and prescribing decisions in primary care. They also note that data showing that non-Hispanic black patients had shorter visits than non-Hispanic white patients seeing the same physician, which could result in accumulation of time disparities that can potentially contribute to racial disparities. They conclude that the data “should motivate organizations and policy makers to detect, interrogate, and address underlying systemic causes such as structural racism.”
It would be interesting to compare data pulled from Athenahealth users and that from users of other EHRs that may have varying levels of clinical decision support or guidelines content within the clinical workflows. In my community, the user base of the Athenahealth EHR tends more towards an independent primary care practice user base. Practices that are owned by or affiliated with the large health systems or academic institutions tend to use a different EHR, as they do across the US. Therefore, using data from one vendor alone might not be representative of primary care practices across the US.
It would also be interesting to control the data for owned versus independent practices, large versus small, and those who are participating in risk-based contracts versus those who aren’t. I’ve found that certain kinds of practices tend to have a more systems-based approach that can make short appointments more efficient than they might be elsewhere.
I work with physicians who practice in a face-to-face environment, those who practice entirely via telehealth, and those who either do a hybrid approach from within their practices or who practice at separate in-person and telehealth jobs. I’ve seen telehealth physicians held to standards that some of their in-person counterparts aren’t monitored for, because there’s a suspicion that somehow telehealth physicians are doing a worse job at following guidelines and standards than their in-person colleagues.
It would also be interesting to compare and contrast the data for telehealth visits done by third-party providers versus those delivered by the patient’s medical home. You would also have to look at hybridized care models such as a primary care office that uses an acute care telehealth pool that’s part of an overall health system, or primary care offices that allow third-party providers to work within their own EHR.
There’s not a tremendous body of literature looking at the length of telehealth visits compared to the outcomes of those visits, and maybe someday I can be part of the research into how telehealth can best be used for what kinds of care and what clinical decision tools work best to provide care in different environments. It’s been a long time since I was involved in research, but I enjoyed it. I’ve just entered a new Maintenance of Certification cycle with my specialty board and a practice improvement project is in order, so one never knows.
What do you think about the association with visit length and care quality? What have your experiences been from the patient side? Leave a comment or email me.
Email Dr. Jayne.