Lifting Coronavirus Disease-19 Shelter-in-Place Restrictions: Impact on Heart Failure Hospitalizations in Northeast Georgia

To the Editor:

During the coronavirus disease-2019 (COVID-19) pandemic, the heart failure (HF) community witnessed unprecedented decreases in HF-related patient visits and hospitalizations. Studies to date have focused on the impact of the initial phase of the COVID-19 lockdown of HF hospitalization trends across multiple countries.1, 2, 3, 4 It is likely that patients with HF, who are at higher risk for morbidity and mortality from COVID-19, delay seeking care in large part owing to anxiety about contracting the virus. Nevertheless, the concerns may resolve over time with availability of treatment for COVID-19 and overwhelming HF symptom burden. Furthermore, the decision to seek acute HF care may be affected by the announcements of state governments relaxing shelter-in-place restrictions and allowing opening of businesses.

On March 14, 2020, the state of Georgia declared a public health state of emergency, followed by an order of shelter in place being effective on April 3. On April 24, Georgia became the first state to allow businesses to reopen after the coronavirus shutdown. Here, we describe the trend in HF hospitalizations before the state of emergency, during the shelter-in-place order, and after the reopening. Data were collected at the Northeast Georgia Health System, a large, quaternary care, community health system with 713 beds and a large HF/left ventricular assist device program.

Methods

We retrospectively collected the number of weekly HF hospitalizations using the electronic health record system from February 1 to June 12, 2020. Hospitalizations primarily owing to HF were identified using diagnosis-related group codes 291, 292, and 293. These data were compared to the same period in 2019 (January 31 [owing to a shorter February] to June 12).

Categorical data were summarized using percentages, while numerical data were summarized using means and standard deviations. Differences in means were tested using the independent t test. The assumption of normality was verified with Shapiro-Wilk´s test and homogeneity of variance with Levene test. Differences in proportions were tested using 2-tailed χ² test. Statistical significance was assessed at the 0.05 level. Statistical analysis was performed using MedCalc Statistical Software (Version 19.4, 2020, MedCalc Software Ltd. Ostend, Belgium).

Results

There were 578 and 480 HF hospitalizations during the periods of interest in 2019 and 2020, respectively. No significant difference between the 2 cohorts in terms of age or the percentage of females was observed before the state of emergency (age, P = .497; female %, P = .055), during the shelter-in-place order (age, P = .083; female %, P = .196), or after the reopening (age, P = .083; female %, P = .507). There was a significant decrease in the percentage of white patients seen after the declaration of state of emergency, with a 9% and 13% decrease in weeks 6 to 14 and weeks 15 to 19, respectively. Length of stay did not change in weeks 1 to 14 (n = 5.56 ± 5.11 days vs n = 5.69 ± 4.55 days, P = .712). However, in weeks 15 to 19, the length of stay in 2020 was shorter than in 2019 (n = 4.9 ± 3.4 days vs n = 6.0 ± 4.3 days, P = .012).

Weekly hospitalizations for HF during weeks 1 to 5 were comparable between 2019 and 2020, with average weekly admission rates of n = 27.4 and n = 28.2 (P = .785), respectively (Fig. 1 ). However, from weeks 6 to 14, a significant decrease in hospitalizations was observed accompanied by an increase of confirmed COVID-19 cases, compared with the weekly admission rate in 2019. The mean weekly admission rate was reduced by 36.5% from 33.2 ± 4.7 in 2019 to 21.1 ± 5.3 in 2020 (P < .001). The lowest weekly hospitalizations were found 1 week before the order of shelter in place; afterward, the admission rate rebounded continuously. From weeks 15 to 19, 2 weeks after the reopening, weekly hospitalizations returned to a similar level observed in 2019 (28.4 ± 4.2 days in 2019 vs 29.8 ± 5.0 days in 2020, P = .288).

Fig. 1.

Heart failure hospitalizations during the coronavirus disease-2019 (COVID-19) pandemic in the Northeast Georgia Health System (NGHS).

Discussion

Our findings echo those of previous studies, with a similar pattern of fewer hospitalizations and a disproportionate effect on minority racial groups. Unique to our analysis is that Georgia was the first state to lift shelter-in-place orders. We observed that HF hospitalization rates returned to historical values 2 weeks after reopening and was maintained thereafter. No surge in HF hospitalization was seen, despite the “hospitalization debt” incurred during the public health state of emergency and shelter-in-place periods. The decreased length of stay likely reflects patient preference for earlier discharge and physician response in an effort to minimize potential exposure to COVID-19. As states lift shelter-in-place orders, healthcare facilities should be prepared for an increase in HF hospitalizations. It is unclear if a surge in admissions will be seen at a future date. At the time of this writing, COVID-19 cases are spiking in southern states including Georgia. Whether the recovered HF hospitalization in our health system will be durable amid this alarming increase is unknown.