Characterization of Prolonged COVID-19 Symptoms in an Outpatient Telemedicine Clinic.

We identified patients with coronavirus disease 2019 in a telemedicine clinic who requested ongoing follow-up calls 6 weeks after symptom onset. In this group, respiratory symptoms are the most common complaints, asthma and lung disease are frequent comorbidities, and patients often have not returned to work or usual activity.

While the clinical course of ambulatory patients with coronavirus disease 2019 (COVID-19) has been described in early reports [1-4], persistent symptoms beyond 5 weeks are not yet well characterized. Data illustrating delayed recovery are available from the severe acute respiratory virus disease (SARS) outbreak in 2003 [5] as well as community-acquired pneumonia [6], but these reports are limited to post-hospitalization cohorts. Anecdotally, some patients with COVID-19 report delayed return to routine activity, even with mild illness.

The Emory Clinic Virtual Outpatient Management Clinic (VOMC) is a telemedicine program for the care of adults with COVID-19 during isolation at home. The VOMC follows patients with confirmed COVID-19 with regular telephone calls for a duration of 7–21 days depending on patient symptom severity, comorbidities, and age [7]. The calls may be continued beyond 21 days for ongoing symptom monitoring at the request of the patient and/or provider (registered nurse or advanced practice provider). We seek to describe the persistent symptoms experienced by patients with mild COVID-19 by reviewing records of those who requested follow-up VOMC care for more than the planned 21 days and more than 6 weeks beyond symptom onset. We hypothesize that specific comorbidities may be associated with prolonged symptom duration in comparison with the overall VOMC population.

METHODS

We conducted a search of the patients enrolled in the VOMC between March 24 and May 26, 2020, who received their final VOMC follow-up call >6 weeks after the date of symptom onset. Exclusion criteria were (1) unclear onset date, (2) VOMC follow-up call duration of <3 weeks, and (3) hospitalization before entering the VOMC. Charts were reviewed through June 8, at which time all eligible patients had been discharged from the VOMC.

Chart review included (1) verification of patient demographics and comorbidities documented at the VOMC intake visit, (2) verification of symptom onset dates, (3) review of follow-up notes during the sixth week of symptoms, (4) review of return to work advice/disability letters, and (5) review of final notes for health status at time of VOMC discharge. We obtained data from standardized lists of symptoms and comorbidities coded in the VOMC notes but allowed for specified “other” symptoms and comorbidities specifically denoted in the provider narrative portion(s) of the note to be coded as “other” for later analysis. For patients who required additional medical evaluation after the acute period (defined as an in-person or telemedicine visit at least 3 weeks into illness), we reviewed evaluation notes, diagnostics, and final diagnoses (including “alternate diagnoses” and “contributing diagnoses” based on provider documentation). Delayed return to activity was coded as present only if a provider note during the sixth week of symptoms specifically noted less ability to perform physical activity (eg, walk, jog, or run errands) compared with activity status immediately before illness. All chart review data were entered into a standardized template capturing all elements that we have reported in results.

Data for specific comorbidities were available for the overall VOMC cohort [7], extracted by data pull and available for comparison with the comorbidities for our “prolonged symptom” subset in this study. For the comparison of comorbidities, we used the same data source (intake note) for the overall cohort and the prolonged symptom subset. We included additional comorbidities in our chart review template if identified in 2 or more charts (prolonged group only). Results were analyzed in Microsoft Excel using descriptive statistics.

RESULTS

A total of 551 patients were identified as being monitored by VOMC during the study period. Of these, 496 were confirmed to have COVID-19 by nasopharyngeal polymerase chain reaction and enrolled in VOMC care. We identified 51 (9.4%) as receiving calls >6 weeks after symptom onset and arrived at a total of 26 (4.8%) “prolonged cases” after exclusions: We removed 5 patients due to unclear start dates (3 reporting symptom onset >2 months before local transmission was suspected, 2 with negative testing at symptom onset and retesting positive >1 month later), 10 patients due to total symptoms <6 weeks (dates clarified in chart review), and 10 patients due to receiving <21 days of VOMC calls (eg, an initial VOMC visit in the fifth week of illness, followed for 1 week and discharged).

Table 1 outlines the demographics and comorbidities of patients in the “prolonged cohort” compared with the overall cohort. The majority of patients were female (76.9%), and the median age (range) was 47.5 (23–78) years. Racial demographics were as follows: 14 (53.8%) African American, 5 (19.2%) Caucasian, 7 (26.9%) other or not recorded. The most common conditions noted were BMI >30 (53.8%), asthma (42.3%), allergies (34.6%), and hypertension (34.6%).

Table 1.

Demographics and Comorbid Conditions

	Prolongeda (n = 26), No. (%)	Nonprolonged (n = 470), No. (%)
Age category
18–29 y	2 (7.7)	76 (16.2)
30–39 y	5 (19.2)	79 (16.8)
40–49 y	7 (26.9)	99 (21.1)
50–59 y	7 (26.9)	108 (23.0)
60–69 y	2 (7.7)	82 (17.4)
≥70 y	3 (11.5)	26 (5.5)
Sex
Female	20 (76.9)	310 (66.0)
Male	6 (23.1)	160 (34.0)
Race
Black	14 (53.8)	238 (50.6)
White	5 (19.2)	92 (19.6)
Other/not recorded	7 (26.9)	140 (29.8)
Standard comorbidities from VOMC note
No comorbidity	6 (23.1)b	174 (37.0)
Age >60 y	5 (19.2)	108 (23.0)
BMI >30	14 (53.8)	196 (41.7)
Hypertension	9 (34.6)	156 (33.2)
Coronary artery disease	1 (3.8)	19 (4.0)
Diabetes	1 (3.8)	65 (13.8)
Immunosuppressionc	1 (3.8)	26 (5.5)
Chronic kidney disease	1 (3.8)	14 (3.0)
Lung disease	3 (11.5)	11 (2.3)
Asthma	10 (42.3)	59 (12.6)
Additional comorbiditiesd
Allergies	9 (34.6)	NR
Mood disorder	6 (23.1)	NR
Migraines	4 (15.4)	NR
Autoimmune disorder	4 (15.4)	NR
Previous tobacco usee	5 (19.2)	NR

Abbreviations: BMI, body mass index; NR, not recorded in standard VOMC template; VOMC, Emory Clinic Virtual Outpatient Management Clinic.

aProlonged was defined as patients who continue VOMC for >3 weeks, extending >6 weeks beyond symptom onset. Nonprolonged is all other patients in the cohort, presented for comparison (note that the baseline characteristics of this cohort are described in O’Keefe et al. [7]).

bWhen including the “other” comorbidites identified in chart review (denoted d in the table), the number is 1 (3.8%).

cUse of biologic or other immunosuppressive medications including chronic corticosteroid at ≥20 mg prednisone daily, detectable HIV viral load, or CD4 count <200 cells/mm3.

dThese comorbidities were coded from the “other” category on chart review of the prolonged cohort.

eNo current smokers were identified in the prolonged cohort.

Patients with persistent symptoms entered the VOMC a median (range) of 9.5 (4–39) days after symptom onset and were followed by the VOMC for a median (range) of 38 (21–49) days. The time from symptom onset to discharge from the VOMC was a median (range) of 47.5 (42–80) days. At VOMC discharge, 24 (92.3%) patients reported significant improvement in symptoms, with only 7 (26.9%) reporting that they were at baseline health (symptom free).

The presence of symptoms in week 6 is presented in Table 2. Respiratory symptoms were most common, reported in 23 patients (88.5%), most frequently cough, shortness of breath with exertion, sinus congestion, and chest tightness. Other common symptoms include fatigue (17 patients, 65%) and headache (13 patients, 50%). Less commonly reported, 9 patients (34.6%) had persistent gastrointestinal symptoms, 6 patients (23%) complained of palpitations, and 3 had persistent low-grade fevers. Of note, 18 (69.2%) reported at least 4 concurrent symptoms.

Table 2.

Symptoms Reported by Patients in Sixth Week

	No. (%)
General	20 (76.9)
Fever	3 (11.5)
Chills	3 (11.5)
Body aches	7 (26.9)
Fatiguea	17 (65.4)
Weaknessa	7 (26.9)
Joint pain	8 (30.8)
Sweatsa	2 (7.7)
Neurologic	23 (88.5)
Loss of smell	8 (30.8)
Altered tastea	1 (3.8)
Confusion	1 (3.8)
Dizziness	3 (11.5)
Headache	13 (50)
Memory lossa	4 (15.4)
Sleep disturbancea	7 (26.9)
Anxietya	8 (30.8)
Respiratory	23 (88.5)
Sore throat	5 (19.2)
Sinus congestion	12 (46.2)
Cough	14 (53.8)
Shortness of breath at rest	3 (11.5)
Shortness of breath with exertion	13 (50)
Wheezing	3 (11.5)
Chest tightness	10 (42.3)
Gastrointestinal	9 (34.6)
Diarrhea	3 (11.5)
Abdominal pain	1 (3.8)
Nausea	5 (19.2)
Anorexiaa	3 (11.5)
Other
Rash	1 (3.8)
Palpitationsa	6 (23.1)
Total number of symptoms
1–3	7 (26.9)
4–6	13 (50)
7–9	3 (11.5)
10 or more	2 (7.7)

Symptoms are grouped by organ systems, noted in bold, which include the total number of patients with any symptoms from within the category.

aThese symptoms identified from “other” category on chart review.

Due to the presence of persistent symptoms, 16 (61.5%) patients delayed return to work at least 5 weeks from symptom onset and 17 (65.3%) delayed return to activity. The most common reasons cited for delayed return to work or activity were fatigue and weakness.

The majority of patients (73%) underwent further evaluation at 1 or more sites at least 3 weeks after symptom onset, either in the emergency room (n = 5, 26.3%), respiratory clinic (n = 12, 63%), or in a telemedicine visit with specialist (n = 4, 21%) or primary care provider (n = 4, 21%). Common tests included chest x-ray (n = 10), chest computed tomography (n = 7), labs (n = 8), and echocardiogram (n = 4). An alternate non-COVID-19 diagnosis was reached only for 1 patient (exacerbation of heart failure). A contributing diagnosis (to delayed improvement in COVID-19) was suspected for 13 patients (50%), most commonly allergic rhinitis (n = 7, 58.3%), followed by asthma (n = 5, 41.7%) and bronchiectasis in 2 patients (16.7%).

DISCUSSION

Our data demonstrate that a subset of outpatients with mild COVID-19 will experience persistent symptoms, here defined as a period >6 weeks. While the majority of patients identified in this study experienced impairment that limited usual activity, we found that most patients (92.3%) had an improving symptom course in the sixth week of symptoms, though long-term duration of symptoms is still unknown. A smaller proportion (26.9%) reported a return to baseline health by the time of discharge from the VOMC.

In this cohort, the majority of patients sought further evaluation for their symptoms, with only 1 patient identified to have a non-COVID-19 alternate diagnosis. Importantly, in patients evaluated for persistent symptoms, a contributing atopic diagnosis or chronic lung disease was often suspected, which led to specific directed treatments. This possible association is also suggested by comparison of comorbidities between the prolonged symptom group and the overall VOMC cohort (Table 1); asthma and chronic lung disease (prospectively coded at intake visit) appeared more frequently in the patients identified in the persistent symptom cohort for this study. Anecdotally, we note that many providers reported that inhaled bronchodilators and corticosteroids were effective for prolonged COVID-19 symptoms (https://youtu.be/ecPjhdVf41k), and our findings in this report strengthen the case for further research.

While respiratory symptoms (cough, dyspnea on exertion, and other) are most common, a variety of symptoms may present across organ systems including neurologic, cardiac, and gastrointestinal. These manifestations merit further investigation as possible evidence of organ-specific dysfunction caused by “mild” COVID-19 in outpatients.

Limitations

Our data represent a specific population of patients who enrolled in a telemedicine program at a single center and may not be generalizable to other populations. Additionally, the request for ongoing follow-up calls was not standardized across the cohort, so we cannot be certain which symptoms or functional status concerns prompted the continuation of care. It is therefore likely that our data do not capture all patients with persistent symptoms. In a separate phone call survey project, 4/148 (2.7%) nonhospitalized patients who were called after VOMC discharge (median symptom day, 63) reported significant symptoms requiring ongoing medical care by primary care providers [8].

The study period (March–May 2020) overlapped with the spring pollen season in Georgia, which could explain the occurrence of contributing atopic diagnoses in this study. The peak pollen counts, however, occurred from March 20 to April 10, and the dates of discharge of our prolonged symptom cohort from the VOMC were April 24 to June 8 (median date, May 18); the median dates of the sixth week of symptoms used for our analysis were May 4 to May 11. Furthermore, the provider documentation used in our analysis attributed the symptoms to COVID-19 as the primary active diagnosis in all care plans except for a single “alternate diagnosis” case.

CONCLUSIONS

For a subset of patients with COVID-19 (4.8% in our cohort), symptom duration lasts >5 weeks and impacts their ability to return to work and activity. The most common persistent symptoms are respiratory in nature. These patients may be more likely to have underlying allergic and lung conditions than the general telemedicine VOMC follow-up cohort. Further research is needed to determine the long-term effects of COVID-19 in patients with persistent symptoms.