COVID-19 Impact on Intern Exposure to Common Inpatient Diagnoses

Our study was conducted at a freestanding tertiary care pediatric hospital with a large (>50 interns per year) pediatric residency program. Categorical pediatric residents typically do 3 4-week blocks on the PHM service in their intern year, followed by 1 supervisory block during their third year.

The PHM service at our institution is composed of 4 teams, each with 1 attending, 1 upper-level resident, 2 to 4 interns, 1 to 2 medical students, and, intermittently, a PHM fellow.

We analyzed electronic medical record data to compare hospitalizations from 8 4-week academic blocks from June 24, 2020, through February 2, 2021, with the same academic blocks from 2017 to 2020. We used a previously built educational dashboard at our institution, in which International Classification of Disease, Tenth Revision, (ICD-10) diagnostic codes are assigned to PHM core competency diagnoses and resident exposures to these diagnoses before and during the pandemic are compared.^6,8  Multiple ICD-10 codes are assigned to each core competency diagnosis (eg, 6 ICD-10 codes are assigned to bronchiolitis and 18 ICD-10 codes are assigned to asthma). We used previously published PHM core competencies endorsed by the Section on Hospital Medicine and Academic Pediatric Association, with slight modifications for patient populations cared for at our institution.⁶ 

We attributed the patient’s diagnoses to an intern if they wrote a history and physical note, progress note, or discharge summary for the patient during that patient encounter. Once a patient was attributed to a resident on the basis of note authorship, all associated ICD-10 codes at the time of discharge were assigned to the appropriate core competency diagnosis by our automated dashboard. If a patient had multiple core competency diagnoses (eg, respiratory distress and pneumonia), they were counted for each diagnosis. If an intern wrote multiple notes on the same patient, each note was counted. In doing so, we were able to track resident exposure to each core competency diagnosis, as well as overall number of notes written each month. We only included notes written by interns rotating on their PHM academic block. This means that notes written by upper-level residents or “cross-cover” residents covering for a weekend were excluded. To quantify shifts worked per intern, we counted a shift worked as any date that an intern wrote at least 1 note during their PHM block.

Our institutional review board approved this study.

We compared the median number of notes written per intern per block overall, as well as for each core competency diagnosis, and the median number of shifts worked per intern per block before and during the COVID-19 pandemic. We used the Mann-Whitney U test to compare the medians. Because annual and seasonal variations in hospital admissions exist irrespective of COVID, we intentionally chose a P value <.001 as being considered statistically significant.

We identified 10 627 hospitalizations in academic blocks 1 to 8 during the 3-year pre–COVID-19 group and 2750 hospitalizations in the same academic blocks in the COVID-19 pandemic group. Median counts of notes written per intern per block were significantly reduced in the COVID-19 group compared with the pre–COVID-19 group (96 [interquartile range (IQR): 81–119] vs 129 [IQR: 110–160]; P < .001). Median counts of intern notes per academic block were significantly reduced for every academic block except block 8 (Fig 1). Additionally, the median number of notes for many common inpatient diagnoses, most notably respiratory illnesses and certain infections, were significantly reduced (Fig 2; Table 1) but were higher for mental health (4 [IQR: 2–9] vs 2 [IQR: 1–6]; P < .001) and suicidality (4.5 [IQR: 2–8] vs 0 [IQR: 0–2]; P < .001). Median shifts worked per intern per block were reduced by ∼1 shift in the COVID-19 group compared with the pre–COVID-19 group (22 [IQR: 21–23] vs 23 [IQR: 22–24]; P < .001).

Our institution experienced a decrease in hospitalizations for many common pediatric diagnoses, which mirrors the global experience.^2,6  The most dramatic effect is seen in admissions for respiratory complaints like asthma, bronchiolitis, pneumonia, and respiratory distress. However, nonrespiratory conditions have also been impacted, including neonatal fever, head and neck infections (eg, pharyngitis, croup, and sinusitis), skin and soft tissue infections, and Kawasaki disease. Admissions for mental health and suicidality increased abruptly, confirming previous findings.⁹ 

This decrease in pediatric hospitalizations has been reported previously,^2,10  with a recently published study describing a decrease in pediatric interns’ clinical exposures during the COVID-19 pandemic surge (March 2020 to June 2020).¹¹  With our study, we confirm their findings and further describe the persistence of this significant decline in clinical exposure from July 2020 to February 2021, which was previously not examined. Additionally, because we used an educational dashboard that incorporated established PHM core competency diagnoses, we were able to quantify the effect of decreased hospital census on pediatric interns’ exposure to specific clinical diagnoses that we find meaningful from an educational standpoint. Finally, we believe our study to be the first in which the striking increase in pediatric intern exposure to patients admitted for suicidality is reported.

Our results may have broad implications for pediatric residency training. Although we recognize that pediatric resident education is not solely dependent on high patient volume and exposure, resident education has traditionally emphasized the importance of clinical experiential learning.^12–14  Therefore, the decreased patient volumes and significantly fewer opportunities for clinical experiential learning encountered by the current intern class may lead to gaps in their training that need to be addressed. In many residency programs, the majority of PHM rotations are during intern year, but this intern cohort may need additional clinical time in the emergency department or inpatient wards in the future to enhance their exposure to common inpatient diagnoses. Additionally, future study would be needed to definitively describe the impact the COVID-19 pandemic has had on the objective competence of the affected pediatric trainee cohorts.

Residency programs could also consider other curricular innovations, such as additional trainings, patient simulations, or didactic curriculum, to provide increased educational exposure to many of these diagnoses. Patient simulations may be particularly important in helping develop clinical recognition and practice in managing respiratory distress and other urgent inpatient clinical conditions. Pediatric faculty at teaching hospitals should also be aware of the potential need for increased supervision and support in these clinical situations as the current intern class becomes supervisory residents.

The dramatic increase in admissions for suicidal ideation or attempts may have effects on resident well-being. Programs should consider increased monitoring or frequent check-ins for pediatric resident burnout during this already stressful time.^17,18  Given the national pediatric mental health crisis we are currently facing,⁹  pediatric residency programs may consider additional training for residents with regard to deescalation techniques and commonly used pediatric psychiatric medications to maximize the educational potential and resident comfort level in caring for these patients.

Finally, we did find that pediatric interns worked ∼1 less shift per block during the pandemic. There were no programmatic changes to the intern’s cross-cover or PHM rotation schedules, and no changes to the note-writing expectations of pediatric interns during our study period. Notably, at our institution, if an intern misses a shift because of illness, they do not make up the day later in the year. Although we believe that the reduction in number of shifts worked per block contributes to our findings above, 1 less shift per block would seem an inadequate explanation for the profound decreases in clinical exposure that we saw during the COVID-19 pandemic.

This single-center study may limit generalizability. The existing educational dashboard only captures resident notes written on the PHM rotation and may not reflect the full residency experience, although similar decreased prevalence of hospitalizations and respiratory illnesses has been noted across the continuum of pediatrics.¹  Finally, although annual variations in hospitalization rates may exist, a lower P value was intentionally chosen to denote statistical significance to account for this potential inherent variation.

As a result of the COVID-19 pandemic, pediatric hospitalization rates and respiratory illness prevalence have dramatically decreased, and our findings reveal a significant reduction in intern exposure to many common inpatient pediatric diagnoses. Residency programs should consider innovative curricular modifications to ensure adequate knowledge and experience for residents affected by this pandemic.