Characterization and clinical outcomes of outpatients with subacute or chronic post COVID-19 cough: a real-world study

Study population

We collected data on patients admitted to our outpatient unit between August 2023 and February 2024 with a diagnosis of cough who met all the following criteria: (1) aged between 16 and 65 years; (2) cough following a documented COVID-19 infection, confirmed by a nucleic acid test or antigen test; (3) cough duration of at least 3 weeks; and (4) absence of fever.

To ensure that the cough could be attributed to a COVID-19 infection, patients with any of the following criteria were excluded: (1) recent radiographic evidence (X-ray or CT within 3 days) of pneumonia or pulmonary fibrosis; (2) suspected gastroesophageal reflux disease (GERD)-related cough (e.g., acid reflux, heartburn) or upper airway cough syndrome (UACS) (e.g., significant postnasal drip); (3) concurrent influenza virus infection, confirmed by a nucleic acid test or antigen test; (4) cough induced by angiotensin-converting enzyme inhibitors (ACEI), which resolved upon cessation of the drug; (5) lack of consent; or (6) asthma with poorly controlled symptoms.

Clinical data collection and effectiveness assessment

The following data were collected: (1) demographic and baseline information, including age, gender, smoking history, vaccination status, comorbidities (such as hypertension and diabetes), and concurrent symptoms (e.g., expectoration and abnormal laryngeal sensations); (2) blood tests and equipment assessments, when available, including C-reactive protein (CRP), white blood cell (WBC) count, and fractional exhaled nitric oxide (FeNO) levels; and (3) prescription details and patient-reported cough severity during the initial visit and subsequent follow-up appointments.

Follow-up assessments were conducted via phone every 2 weeks following the patients’ initial visit, with cough severity measured using the Visual Analog Scale (VAS), graded from 0 to 100. A VAS score of 70 indicated that the cough significantly impacted usual activities, such as sleep or talking. If a patient reported cessation of coughing during the first follow-up, the assessment was concluded; otherwise, a second follow-up was scheduled for 2 weeks later.

The efficacy of the treatment regimen was evaluated based on changes in VAS scores. A regimen was considered non-responsive (NR) if a patient’s VAS score remained unchanged or increased. A complete response (CR) was defined as a decrease in VAS score by over 30 points or reaching a final VAS score of 0. Otherwise, the regimen was categorized as a partial response (PR). The final follow-up was conducted on March 7th, 2024.

Statistical analysis

Continuous data were described as the mean ± standard deviation (SD) or median (interquartile range (IQR)) and were analyzed using t-tests or Wilcoxon signed-rank tests, depending on the data distribution. Categorical data were expressed as the number of cases and proportions, with the Chi-square test applied for hypothesis testing. Logistic regression analysis was employed to identify risk factors associated with cough relief (defined as a VAS decrease of 30 points or more). Variables with a p-value ≤ 0.05 in univariate logistic regression were further examined using multivariate logistic regression models. All statistical analyses and visualizations were performed using R (version 4.2.3; R Core Team, 2023) or GraphPad Prism (version 8.4.0). A p-value of < 0.05 was considered statistically significant.

Patient characteristics

A total of 4,102 patients were initially screened, of whom 192 were selected for further assessment. After excluding 51 patients, 141 individuals were ultimately included in the study.

As shown in Table 1, the patients had a median age of 42 years (mean ± SD: 42 ± 14), with females accounting for 70% of the cohort. The majority (117, 83%) had been vaccinated once (4, 3.4%) or more(113, 96.6%). A minority were smokers or ex-smokers (17, 12%).

The patients with subacute cough accounted for 68.1% of the population, and the median duration of cough was 8 weeks (interquartile range: 4–12 weeks). Dry cough was the only symptom reported by 85 patients (60%) while the other type was a cough with sputum, usually whitish. Additionally, over half of the patients (76, 54%) experienced abnormal laryngeal sensations, such as a dry or itchy throat. Diurnal variation in coughing was insignificant for 32% of patients, while 48% experienced nocturnal coughing, and 21% coughed during the daytime.

Forty-one patients completed the fractional exhaled nitric oxide (FeNO) test, with 31 results within normal range and 10 cases (24%) exhibiting elevated levels. Routine blood tests were performed on 17 patients, and C-reactive protein (CRP) tests were conducted on 15 patients, both of which showed normal levels of white blood cells (WBC) and CRP.

Treatment and outcome

Among the patients, 131 were prescribed oral medications. Compound methoxyphenamine (CMP) capsules were the most frequently prescribed, followed by inhaled corticosteroids/long-acting beta-2 agonist (ICS/LABA), montelukast sodium tablets, and Chinese patent drugs (CPD), including Suhuang Zhike capsules and Compound Loquat Dew. Ten patients opted not to receive any medications based on their personal preference, given their normal auxiliary test results.

Throughout the follow-up period, patients who received prescribed medications demonstrated significant improvement. The median initial Visual Analog Scale (VAS) score for this group was 60 (interquartile range: 50–70), which decreased to 30 (interquartile range: 10–50) at 2 weeks, and further to 20 (interquartile range: 0–40) at 4 weeks. The overall response rate (ORR), defined as the sum of the complete response rate and partial response rate, was 86%. Among patients who opted not to take medications, the median initial VAS score was 53 (interquartile range: 50–55), decreasing to 30 (interquartile range: 30–40) at 2 weeks, and 30 (interquartile range: 21–38) at 4 weeks (as shown in Fig. 2). The natural overall response rate for this group was 80%. Patients who received prescribed medications showed greater improvement in VAS scores compared to those who did not, both at the 2-week time point (26.5 ± 20.8 vs. 15.5 ± 15.1, p = 0.054) and at 4 weeks (36.9 ± 24.4 vs. 21.0 ± 18.4, p = 0.025). We also compared the primary VAS scores of those who received vaccination (n = 117) with those who did not (n = 24), and the results indicated that those who received vaccination had a slightly lower primary VAS, but the difference was not significant (56.41 ± 16.77 vs. 58 ± 13.92, p = 0.47).

To provide a more comprehensive analysis of the response to prescribed medications, patients were categorized into four groups based on the medication they were prescribed: the CMP group, montelukast group, ICS/LABA group, and CPD group. The CMP group had the largest number of patients (n = 81), with six patients exhibiting a partial response and 61 achieving a complete response. The ICS/LABA group, consisting of 46 patients, showed the highest complete response rate (CRR) of 87.0% (n = 40) and an overall response rate (ORR) of 95.7%. In the montelukast group, which included 46 patients, eight had a partial response and 30 had a complete response. The CPD group comprised 30 patients, of whom eight achieved a partial response and 16 a complete response (as shown in Fig. 3).

Logistic regression analysis was performed to identify factors associated with cough relief. The results indicated that patients with a cough duration of no more than 8 weeks (OR = 3.73, 95% CI [1.75–7.97], p < 0.001), those treated with ICS/LABA (OR = 4.44, 95% CI [1.72–11.51], p = 0.002), or those not treated with CPD (OR = 0.42, 95% CI [0.18–0.97], p = 0.043) were more likely to experience cough relief at 4 weeks. Multivariate logistic regression analysis, adjusted for age and gender, further confirmed that a cough duration of no more than 8 weeks (OR = 4.28, 95% CI [1.90–10.0], p < 0.001) and treatment with ICS/LABA (OR = 4.42, 95% CI [1.70–13.3], p = 0.004) were independently associated with cough relief (as shown in Fig. 4).

To elucidate the effectiveness of specific regimens, patients were categorized based on their detailed prescriptions. The detailed results are shown in Table 2. We listed the top six most frequent regimens for clinical reference. The regimen comprising CMP Capsules and montelukast was the most common, involving 27 patients, with an ORR of 70.4% and a CRR of 77.8%. This was followed by the regimen of CMP Capsules alone, which involved 26 patients and demonstrated an ORR of 76.9% and a CRR of 84.6%. The regimen containing only ICS/LABA or CPD was the third most frequent, involving 19 patients, with ORRs of 100% and 78.9%, and CRRs of 89.5% and 47.4%, respectively. Additionally, the regimens comprising CMP Capsules and ICS/LABA, and CMP Capsules combined with ICS/LABA and montelukast, involved 12 and nine patients, respectively. The ORRs and CRRs were as follows: for CMP Capsules + ICS/LABA, 83.3% and 83.3%, and for CMP Capsules + ICS/LABA + montelukast, 100% and 100%.