General Internists in Pursuit of Diagnostic Excellence in Primary Care: a #ProudtobeGIM Thread That Unites Us All

M aking a correct and timely diagnosis is not only critical to the provision of safe patient care, but central to our identity as general internists. In its absence, preventable harm can occur from delayed, inappropriate, or omitted tests, procedures, and treatments. However, errors in diagnosis are notoriously challenging to study. In 2015, the National Academies of Sciences, Engineering, and Medicine (previously the Institute of Medicine) galvanized renewed focus on diagnostic safety as an integral aspect of health care quality through the publication of Improving Diagnosis in Health Care. Timely diagnosis is essential for certain cancers, where longer time intervals to diagnosis are associated with poorer outcomes. As with many malignancies, colorectal cancer can often present with non-specific signs and symptoms. For example, rectal bleeding is both a harbinger of the malignant (colorectal cancer) and the relatively benign (hemorrhoids), with a positive predictive value for cancer barely approaching 10%. As general internists, we must first rely on our most revered skill, the history and physical, to help separate the signal from the noise. Yet, the reality is that practicing in the organized chaos of the ambulatory clinic rife with competing interests, such as other presenting concerns and documentation requirements, makes this an exceedingly challenging task. In this issue of JGIM, Percac-Lima and colleagues contribute to the growing body of evidence describing how a diagnosis of cancer can be missed or delayed. Analyzing crosssectional medical record review data from 300 adults presenting with rectal bleeding to academic primary care practices, they found that although almost 90% required colonoscopy as indicated by clinical practice guidelines, orders were placed for only 74%. Of those patients, less than 60% ended up having a colonoscopy within a year. The odds of physicians ordering recommended colonoscopies were significantly lower in patients 40–50 years and in patients with additional primary care visits unrelated to rectal bleeding. These findings echo previous work in this area and, while concerning, are not surprising when situated within the longitudinal and fragmented outpatient experience. Making a diagnosis such as colorectal cancer challenges the resilience of our primary care system. Several processes must be sequentially and successfully completed, each of which are vulnerable to individualand system-level breakdowns. For example, physicians may erroneously attribute microcytic anemia to Bchronic disease^ or miss follow-up on a positive fecal occult blood test. Prior studies have established that process failures in the diagnosis of colorectal cancer most commonly occur during the provider–patient clinical encounter, follow-up of test results, and/or closure of the referral loop. Sure enough, colorectal cancer remains high on the list of conditions that are cited in primary care malpractice claims. So how do we, as general internists, move the needle forward on diagnostic safety in primary care? First and foremost, given our skill set in diagnosis, we are well positioned to help develop and lead a more robust infrastructure encouraging diagnostic safety efforts in outpatient care. Admittedly, we will need more administrative, research, and implementation leadership to make this feasible. To our knowledge, few health systems have implemented some form of governance, oversight, and/or accountability framework to support safety in the ambulatory setting. To paraphrase Deming, our current system is perfectly designed to get the results it gets, which is unfortunately a general lack of investment in improvement efforts. Creating such an infrastructure can help foster a culture incentivizing and rewarding behavior aimed at making diagnosis safer. Whether it’s running outpatient morbidity and mortality conferences, leading a quality improvement team on increasing the follow-up rate of abnormal test results, or conducting applied research, the opportunities are plentiful. Second, we need to revisit the heroism classically attributed to making a challenging diagnosis. Morning reports, rounds, and clinicopathologic conferences are known to venerate the infrequent conditions. In fact, most diagnoses that are missed by general internists are not due to rare diseases, but instead common conditions, such as pneumonia, congestive heart failure, and cancer, which can lead to a significant burden of patient harm. We have learned from many studies, including the one from Percac-Lima and colleagues, that faulty data synthesis and an inadequate history and physical are leading contributors to diagnostic error. Unfortunately, our current system can often make these fundamental clinical skills seem Published online February 8, 2018