Abstract
Background:
Uncomplicated episodes of prolonged acute cough are usually viral and self-limited, but despite evidence and recommendations to the contrary, they are often treated with antibiotics.
Methods:
Mixed cross-sectional and prospective observational study of adults 18 years or older presenting to two urgent care centers with a cough of 7 to 56 days as their chief complaint. Factors associated with cough duration and clinical decisions were analyzed by univariate and multivariate logistic regression.
Results:
Of the 125 enrolled patients, 118 (94%) received an antibiotic, 97 (78%) a cough suppressant, 87 (70%) a systemic corticosteroid, and 39 (31%) a chest X-ray (CXR). Longer duration of cough was associated with the presence of self-reported wheezing or noises (adjusted odds ratio 6.29, 95% CI 1.36–29.16) while the presence of both wheezing and crackles on a clinician chest exam was associated with shorter duration (aOR 0.03, 95% CI 0.00–0.27). A clinician was more likely to order a CXR in patients with dyspnea (aOR 3.01, 95% CI 1.21–7.49), less likely to prescribe a systemic corticosteroid in patients with crackles (aOR 0.27, 95% CI 0.09–0.82), and more likely to prescribe a cough suppressant to older patients (1.04 per additional year of age, 95% CI 1.01–1.07).
Conclusions:
Systemic corticosteroids and cough suppressants are being prescribed at high rates in patients with uncomplicated acute cough in the urgent care setting. Additional studies to determine if similar rates are seen in other urgent care centers, or in other contemporary ambulatory settings are warranted.
Keywords: Cough, Clinical management, Urgent care, Corticosteroids
1. Introduction
Cough is a common reason for seeing an ambulatory physician in the United States; approximately 3% of all visits are due to a cough [1]. The most frequent diagnosis for a cough is upper respiratory tract infection, followed by acute bronchitis [2,3]. Only a small percentage of these episodes of cough, about 5%, are due to pneumonia [4]. A systematic review found that an acute cough lasts a mean duration of 18 days after onset (range of 15 to 28 days), while patient expectations for duration are much lower, between seven to nine days [5].
Antibiotics continue to be widely prescribed for uncomplicated episodes of acute cough, particularly in the outpatient setting [6–8]. Up to 75% of patients seen by outpatient clinicians and diagnosed with acute lower respiratory tract infections (LRTI) are given an antibiotic [6,9,10], despite guideline recommendations against it [11,12]. While a multi-center, placebo-controlled, randomized clinical trial showed no reduction in symptom duration or severity for oral corticosteroids in uncomplicated acute respiratory infections [13], informal observationsinpracticesuggest that corticosteroids are increasingly prescribed for acute LRTI in the absence of asthma or chronic obstructive pulmonary disease (COPD) exacerbation. There is also little evidence that cough suppressants are beneficial, although few studies exist on the benefit of cough suppressants in otherwise healthy adults with acute LRTI [14,15].
There has been a significant increase in the utilization of urgent care centers in the last decade, attributed to decreasing numbers of primary care physicians, perceived urgency of the need for care by patients, and demands for convenience [16]. Management of patients with a cough in these settings is difficult; the expectation from a patient is to have a quick visit and many expect to receive a prescription for treatment. This may influence clinicians, as does the lack of an ongoing continuity relationship and reliance on patient satisfaction as a quality measure and as a component of compensation [17–21].
In the outpatient setting, treatment decisions are based mainly on the clinical presentation of the patient in the form of signs, symptoms, recorded vital signs, and rapid point of care tests for influenza and streptococcal pharyngitis. When community-acquired pneumonia (CAP) is suspected, a chest X-ray (CXR) may be ordered, but the frequency of ordering is unknown in urgent care centers.
As a result of the changing patterns in healthcare utilization, the clinical management of prolonged cough in the urgent care setting is understudied. The goal of this study was twofold: to determine the association between demographics, social factors, and clinical presentation with the likelihood of a cough lasting >14 days from onset, and the association between these same factors and treatment decisions.
2. Methods
2.1. Data collection
Adults 18 years and older with a main or chief complaint of a cough of 7 to 56 days duration were recruited at two urgent care centers in or near Athens, Georgia from February 8, 2017 to December 8, 2017. Patients with moderate or severe asthma, chronic obstructive pulmonary disease (COPD), or who were immunodeficient were excluded. The sites are staffed by a group of physicians and nurse practitioners (collectively described as “clinician” for this study) who rotate between several urgent care centers in the area. On a typical shift there were two to four clinicians staffing each urgent care center.
Enrolled participants were surveyed with questions about their demographic information and their baseline signs and symptoms. The clinician was given a form to document their assessment of the patient and treatment plan. The form included observations the clinician made during the chest exam, the diagnosis, any tests that were ordered (e.g. CXR, rapid flu), and prescriptions given to the patient.
Participants were also given a diary to take home to record their symptom duration and severity for up to 14 days after their visit to the urgent care center. The diary used a 5-point Likert-scale: 0-absent, 1-slightly a problem, 2-moderately bad, 3-bad, 4-very bad (with the exception of post-tussive vomiting and paroxysmal cough, which were simply yes or no). Participants were asked to record their symptoms daily for up to 14 days or until their symptoms resolved.
2.2. Statistical analysis
A univariate and multivariate logistic regression analysis was performed to determine the associations of demographics, social factors, signs, and symptoms with the odds of two different outcomes: 1. Cough >14 days from symptom onset, and 2. Type of clinical management the patient received (CXR, systemic corticosteroid, or cough suppressant). Antibiotic prescriptions were originally included in the analysis plan but ultimately could not be analyzed because 94% of our population received one.
We used univariate logistic regression to determine unadjusted odds ratios (OR) and 95% confidence intervals for each association. Studenťs t-test was used to compare means of continuous variables. Variables were described as statistically significant if the p-value was <0.05 and were noted as a “trend” towards significance if the p-value was between 0.05 and 0.1.
The Akaike Information Criterion (AIC) statistic was used to build a multivariate logistic model. Variables starting with a p-value <0.20 were added until additional variables did not improve the model, defined as a change in AIC of <10 [22–24]. We presented the adjusted odds ratios (aOR) and 95% confidence intervals. Logistic regression analysis was performed using SAS 9.4 (SAS Institute, Cary, NC, USA).
2.3. Ethics approval
The study was approved by the University of Georgia Institutional Review Board for Human Subjects Research (STUDY00003904). A letter of support and permission to recruit patients was received by the medical director of the urgent care centers. Patients provided written consent for inclusion in the study.
3. Results
One hundred and twenty-five patients were enrolled during the recruitment period with a mean age of 41.8 years; 87 (70%) were female, 96 (77%) responded as white, non-Hispanic, and 104 (83%) were diagnosed with acute bronchitis. The mean duration of cough from symptom onset until presenting to the clinic was 15.4 days (95% CI: 13.6–17.2). Ninety-one patients completed follow up and of these, 72 (79%) reported they stopped coughing within the 14 days of follow up.
Among the 72 patients who reported that they stopped coughing during the follow up period, the mean duration from symptom onset to when their cough resolved was 22.4 days (95% CI: 19.6–25.1). Antibiotics were given to 94% (118) of the patients. A CXR was given to 39 patients (31%), systemic corticosteroids were prescribed to 87 patients (70%), and 97 patients (78%) were given a cough suppressant.
3.1. Cough duration from symptom onset
A total of 106 patients were included in the univariate analysis for cough duration from symptom onset (Table 1). Nineteen were excluded due to them not returning at least one diary or survey. Patients that self-reported wheezing or noises when coughing were three times more likely to experience a cough that was longer than 14 days from symptom onset than those with it absent (OR 3.06, 95% CI 1.10–8.45). A respiratory rate of 20 or greater per minute (OR 0.22, 95% CI 0.07–0.70) and wheezes or crackles on a chest exam (OR 0.23, 95% CI 0.08–0.67) decreased likelihood of a prolonged cough when compared to those without.
Table 1.
Unadjusted odds of having a cough duration >14 days from onset (n = 106).
| ORa | 95% CI | p-Value | |
|---|---|---|---|
| Variable | |||
| Demographics and social factors | |||
| Sex | |||
| Female | 1.00 | Reference | |
| Male | 0.55 | (0.19–1.56) | 0.260 |
| Race/ethnicity | |||
| Otherc | 1.00 | Reference | |
| White, non-Hispanic | 0.95 | (0.28–3.22) | 0.940 |
| Income | |||
| >$75k | 1.00 | Reference | |
| <$25k | 1.45 | (0.14–15.16) | 0.252 |
| $25k–$49k | 0.23* | (0.06–0.94) | 0.035 |
| $50k–$74k | 0.33 | (0.07–1.55) | 0.175 |
| Education | |||
| College graduate | 1.00 | Reference | |
| No education | 0.24 | (0.02–3.19) | 0.687 |
| High school graduate | 0.23 | (0.05–1.08) | 0.406 |
| Some college | 0.27 | (0.07–1.06) | 0.550 |
| Current smoker | |||
| Never | 1.00 | Reference | |
| Current | 1.30 | (0.34–4.96) | 0.706 |
| Self-reported signs and symptomsb | |||
| Wheezing | 3.06* | (1.10–8.45) | 0.032 |
| Post-tussive vomiting | 1.97 | (0.60–6.48) | 0.262 |
| Sputum | 1.91 | (0.69–5.32) | 0.216 |
| Paroxysmal cough | 1.90 | (0.63–5.72) | 0.254 |
| Dyspnea | 1.47 | (0.54–4.00) | 0.448 |
| Sneezing | 1.38 | (0.50–3.80) | 0.531 |
| Red or watery eyes | 1.34 | (0.49–3.63) | 0.568 |
| Chills or sweats | 1.28 | (0.47–3.50) | 0.626 |
| Trouble sleeping | 0.91 | (0.18–4.52) | 0.904 |
| Felt warm or feverish | 0.54 | (0.19–1.56) | 0.256 |
| Runny nose | 0.35 | (0.07–1.63) | 0.180 |
| Headache | 0.37 | (0.08–1.73) | 0.207 |
| Clinician recorded signsb | |||
| Heart rate (≥100/min) | 1.02 | (0.26–3.98) | 0.974 |
| Chest sounds (any) | 0.23* | (0.08–0.67) | 0.007 |
| Respiratory rate (≥20/min) | 0.22* | (0.07–0.70) | 0.010 |
| Temperature (>37.7 °C) | 0.21 | (0.01–3.50) | 0.277 |
| Chest exam | |||
| Normal | 1.00 | Reference | |
| Wheezes | 0.55 | (0.12–2.45) | 0.217 |
| Crackles | 0.16 | (0.05–0.58) | 0.245 |
| Wheezes and crackles | 0.07 | (0.01–0.50) | 0.052 |
| Continuous variables | |||
| Heart rate | 1.01 | (0.98–1.05) | 0.507 |
| Age, years | 0.99 | (0.96–1.02) | 0.576 |
| Respiratory rate | 0.89 | (0.71–1.11) | 0.293 |
| Temperature, Celsius | 0.64 | (0.19–2.14) | 0.471 |
CI = confidence interval.
Unadjusted odds ratio: probability modeled patient received a chest x-ray, systemic corticosteroid or cough suppressant.
Reference value is not having sign or symptom.
African-American, Asian, White Hispanic, or Hispanic only (race not selected).
Statistically significant p-value <0.05.
3.2. Ordering a chest X-ray
The unadjusted analysis for predicting CXR, corticosteroid, or a cough suppressant is summarized in Table 2. Patients with a cough for over three weeks were more likely than patients with a shorter cough to have received a CXR (OR 3.03, 95% CI 1.20–7.67). However, when measured as a continuous variable, cough duration was not significantly associated with ordering a CXR. Difficulty breathing (OR 2.64, 95% CI 1.12–6.22) and chills or sweats (OR 2.37, 95% CI 1.08–5.17) increased the likelihood of receiving a CXR compared to when the symptoms were absent. White, non-Hispanic were significantly more likely to receive a CXR than other races and patients of Hispanic origin (OR 3.59, 95% CI 1.15–11.15).
Table 2.
Unadjusted odds of receiving a chest X-ray, or a prescription for a systemic corticosteroid or cough suppressant.
| Chest X-ray |
Systemic corticosteroid |
Cough suppressant |
|||||||
|---|---|---|---|---|---|---|---|---|---|
| ORa | 95% CI | p-Value | ORa | 95% CI | p-Value | ORa | 95% CI | p-Value | |
| Variable | |||||||||
| Demographics and social factors | |||||||||
| Sex | |||||||||
| Female | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| Male | 1.71 | (0.77–3.82) | 0.189 | 1.11 | (0.48–2.55) | 0.816 | 1.81 | (0.67–4.89) | 0.246 |
| Age | |||||||||
| <60 years old | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| ≥60 years old | 2.35 | (0.90–6.13) | 0.080 | 1.11 | (0.40–3.13) | 0.842 | 3.17 | (0.69–14.52) | 0.138 |
| Race | |||||||||
| Otherc | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| White, non-Hispanic | 3.59* | (1.15–11.15) | 0.027 | 0.84 | (0.33–2.11) | 0.707 | 2.84* | (1.14–7.09) | 0.025 |
| Ethnicity | |||||||||
| Non-Hispanic | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| Hispanic | 0.30 | (0.04–2.50) | 0.264 | 3.24 | (0.38–27.27) | 0.280 | 0.45 | (0.10–2.03) | 0.300 |
| Income | |||||||||
| >$75k | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| <$25k | 1.04 | (0.30–3.61) | 0.586 | 1.40 | (0.41–4.74) | 0.679 | 0.38 | (0.11–1.30) | 0.114 |
| $25k–$49k | 1.18 | (0.43–3.22) | 0.763 | 1.23 | (0.47–3.23) | 0.895 | 0.97 | (0.30–3.09) | 0.392 |
| $50k–$74k | 2.36 | (0.86–6.57) | 0.076 | 1.11 | (0.40–3.11) | 0.872 | 0.69 | (0.21–2.24) | 0.941 |
| Education | |||||||||
| College graduate | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| No education | 1.72 | (0.26–11.62) | 0.692 | 0.58 | (0.09–3.92) | 0.620 | 0.65 | (0.06–6.84) | 0.952 |
| High school graduate | 1.58 | (0.58–4.31) | 0.619 | 1.02 | (0.36–2.91) | 0.599 | 0.62 | (0.18–2.16) | 0.984 |
| Some college | 1.06 | (0.43–2.65) | 0.565 | 0.77 | (0.32–1.90) | 0.864 | 0.36 | (0.12–1.03) | 0.170 |
| Current smoker | |||||||||
| Never or former | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| Current | 1.30 | (0.54–3.16) | 0.559 | 1.81 | (0.67–4.89) | 0.246 | 0.83 | (0.31–2.21) | 0.708 |
| Self-reported signs and symptomsb | |||||||||
| Trouble sleeping | 6.62 | (0.77–49.18) | 0.086 | 1.02 | (0.29–3.54) | 0.976 | 0.26 | (0.03–2.11) | 0.209 |
| Cough >21 days | 3.03* | (1.20–7.67) | 0.019 | 0.62 | (0.24–1.58) | 0.316 | 1.46 | (0.45–4.71) | 0.526 |
| Dyspnea | 2.64* | (1.12–6.22) | 0.027 | 0.96 | (0.43–2.10) | 0.908 | 0.90 | (0.38–2.16) | 0.815 |
| Chills or sweats | 2.37* | (1.08–5.17) | 0.031 | 1.54 | (0.72–3.33) | 0.269 | 0.68 | (0.29–1.58) | 0.366 |
| Felt warm or feverish | 2.21 | (0.98–5.01) | 0.056 | 0.79 | (0.36–1.73) | 0.552 | 0.62 | (0.26–1.51) | 0.293 |
| Wheezing or chest sounds | 2.18 | (0.92–5.52) | 0.076 | 1.90 | (0.86–4.17) | 0.110 | 0.57 | (0.22–1.46) | 0.238 |
| Paroxysmal cough | 1.77 | (0.65–4.83) | 0.260 | 0.76 | (0.29–1.98) | 0.569 | 0.54 | (0.17–1.71) | 0.291 |
| Headache | 1.48 | (0.57–3.84) | 0.423 | 0.71 | (0.27–1.85) | 0.482 | 0.21* | (0.05–0.95) | 0.042 |
| Sputum production | 1.37 | (0.57–3.29) | 0.486 | 0.50 | (0.20–1.28) | 0.150 | 0.86 | (0.33–2.27) | 0.767 |
| Runny nose | 1.15 | (0.47–2.79) | 0.764 | 0.59 | (0.23–1.53) | 0.279 | 0.76 | (0.29–2.16) | 0.640 |
| Post-tussive vomiting | 1.09 | (0.49–2.45) | 0.830 | 1.23 | (0.54–2.82) | 0.629 | 1.23 | (0.49–3.10) | 0.659 |
| Red or watery eyes | 0.92 | (0.43–1.98) | 0.838 | 1.35 | (0.63–2.90) | 0.440 | 0.62 | (0.26–1.47) | 0.275 |
| Sneezing | 0.77 | (0.35–1.68) | 0.510 | 1.62 | (0.74–3.53) | 0.226 | 0.62 | (0.25–1.55) | 0.308 |
| Clinician recorded signsb | |||||||||
| Heart rate (≥100/min) | 1.23 | (0.45–3.37) | 0.689 | 4.70* | (1.03–21.37) | 0.046 | 0.84 | (0.28–2.56) | 0.761 |
| Chest sounds (any) | 1.97 | (0.91–4.26) | 0.086 | 0.76 | (0.35–1.67) | 0.492 | 2.11 | (0.82–5.42) | 0.123 |
| Respiratory rate (≥20/min) | 1.35 | (0.49–3.74) | 0.565 | 1.77 | (0.55–5.74) | 0.341 | 1.10 | (0.33–3.62) | 0.879 |
| Temperature (>37.7 °C) | 7.08 | (0.71–70.40) | 0.095 | 1.32 | (0.13–13.13) | 0.812 | 0.27 | (0.04–2.04) | 0.206 |
| Chest sounds | |||||||||
| Normal | 1.00 | Reference | 1.00 | Reference | 1.00 | Reference | |||
| Wheezes | 2.66 | (1.02–6.97) | 0.388 | 2.62* | (0.71–9.70) | 0.014 | 1.33 | (0.44–4.02) | 0.508 |
| Crackles | 1.04 | (0.33–3.24) | 0.212 | 0.35 | (0.13–0.99) | 0.124 | 6.63 | (0.83–52.81) | 0.133 |
| Wheezes and crackles | 4.35 | (0.68–27.94) | 0.230 | 0.26 | (0.04–1.68) | 0.168 | 1.47 | (0.16–13.95) | 0.776 |
| Continuous variables | |||||||||
| Age, years | 1.02 | (0.99–1.04) | 0.178 | 1.01 | (0.98–1.03) | 0.553 | 1.03* | (1.00–1.06) | 0.029 |
| Days coughing | 1.03 | (0.99–1.06) | 0.160 | 0.99 | (0.96–1.03) | 0.575 | 1.01 | (0.97–1.06) | 0.609 |
| Heart rate per minute | 1.01 | (0.99–1.04) | 0.454 | 1.02 | (0.99–1.05) | 0.149 | 0.98 | (0.95–1.00) | 0.095 |
| Respiratory rate per minute | 1.08 | (0.91–1.29) | 0.370 | 0.99 | (0.83–1.19) | 0.967 | 0.98 | (0.80–1.12) | 0.815 |
| Temperature, Celsius | 1.64 | (0.66–4.06) | 0.287 | 1.14 | (0.44–2.95) | 0.782 | 0.51 | (0.19–1.35) | 0.174 |
CI = confidence interval.
Unadjusted odds ratio: probability modeled patient received a chest X-ray, systemic corticosteroid or cough suppressant.
Reference value is not having sign or symptom.
African-American, Asian, White Hispanic, or Hispanic only (race not selected).
Statistically significant p-value <0.05.
3.3. Receiving a prescription
Tachycardia was significantly associated with an increased likelihood of a systemic corticosteroid prescription (OR 4.70, 95% CI 1.03–21.37) while the patient being white, non-Hispanic was significantly associated with receiving a cough suppressant (OR 2.84, 95% CI 1.14–7.09). Patients that received a cough suppressant had a higher mean age of than those who did not (43.5 vs 35.8 years, p = 0.026).
3.4. Multivariate analysis
The multivariate analysis is summarized in Table 3. Patients with self-reported wheezing were significantly more likely to have a cough for longer than 14 days from onset (aOR 6.29, 95% CI 1.36–29.16). Patients were unlikely to experience a prolonged cough when both wheezing and crackles were heard on the chest exam (aOR 0.03, 95% CI 0.00–0.27). Age, heart rate, respiratory rate, and temperature were not associated with cough duration.
Table 3.
Adjusted odds for a cough duration >14 days from onset, or patient was given a chest X-ray, systemic corticosteroid, or cough suppressant.
| Variable | aORa | 95% CI | p-Value |
|---|---|---|---|
| Coughing > 14 days from onset | |||
| Respiratoryrate (≥20/min)b | 0.20 | (0.05–0.72) | 0.014 |
| Self-reported wheezes or noises while coughing | 6.29 | (1.36–29.16) | 0.019 |
| Clinician chest exam | |||
| Normal | 1.00 | Reference | |
| Wheezes | 0.18 | (0.02–1.34) | 0.797 |
| Crackles | 0.12 | (0.03–0.50) | 0.640 |
| Wheezes and crackles | 0.03 | (0.00–0.27) | 0.025 |
| Chest X-ray ordered | |||
| Dyspneab | 3.01 | (1.21–7.49) | 0.018 |
| Cough >21 daysb | 3.08 | (1.16–8.20) | 0.024 |
| Race | |||
| Otherc | 1.00 | Reference | |
| White, non-Hispanic | 3.58 | (1.10–11.66) | 0.034 |
| Prescribed a systemic corticosteroid | |||
| Heart rate (≥100/min)b | 6.10 | (1.22–30.64) | 0.028 |
| Clinician chest exam | |||
| Normal | 1.00 | Reference | |
| Wheezes | 2.59 | (0.69–9.73) | 0.013 |
| Crackles | 0.27 | (0.09–0.82) | 0.051 |
| Wheezes and crackles | 0.31 | (0.05–1.99) | 0.273 |
| Prescribed a cough suppressant | |||
| Headacheb | 0.16 | (0.03–0.75) | 0.020 |
| Age (continuous) | 1.04 | (1.01–1.07) | 0.024 |
| Clinician chest exam | |||
| Normal | 1.00 | Reference | |
| Wheezes, crackles or both | 2.74 | (1.01–7.40) | 0.047 |
CI = confidence interval.
Adjusted odds ratio: probability modeled patient had a cough greater than 14 days or was given a chest X-ray, systemic corticosteroid or cough suppressant.
Reference value is not having sign or symptom.
African-American, Asian, white Hispanic, or Hispanic only (race not selected).
The odds that a CXR was ordered were significantly higher for patients with dyspnea (aOR 3.0, 95% CI 1.2–8.2). A cough duration greater than three weeks at the time of presentation (aOR 3.08, 95% CI 1.16–8.20) and being white, non-Hispanic (aOR 3.58, 95% CI 1.10–11.66) were also independent predictors for receiving a CXR. While significantly associated with obtaining a CXR in the univariate analysis, chills or sweats were not included in the multivariate model.
Patients with a heart rate 100 or greater were more likely to receive a systemic corticosteroid (aOR 6.10, 95% CI 1.22–30.64), while those with crackles on a chest exam were less likely to receive one (aOR 0.27, 95% CI 0.09–0.82). Presence of a headache was associated with a decreased likelihood of a cough suppressant (aOR 0.16, 95% CI 0.03–0.75). When any chest sounds were heard by a clinician (wheezing, crackles, or both), patients were more likely to receive a cough suppressant (aOR 2.74, 95% CI 1.01–7.40). Again, increasing age increased the likelihood that a cough suppressant was prescribed (aOR 1.04, 95% CI 1.01–1.07).
4. Discussion
To the best of our knowledge, this is the first study to describe the prognosis and management decisions of patients with prolonged cough in the urgent care setting. We found several factors associated with the duration of a cough and clinical symptoms that influenced the care the patient received.
Chest sounds, when reported by the patient or heard in the lungs by a clinician during a chest exam, were associated with cough duration. When experienced by a patient, likely as an audible upper respiratory noise (stridor), the likelihood of cough for >2 weeks was increased. When the presence of wheezes and crackles in the lungs was noted on a clinician chest exam, it was associated with a decreased likelihood of prolonged cough.
After performing a chest exam, it may therefore be helpful for physicians and nurse practitioners to ask the patient how they would describe their cough and whether they are experiencing any noises or chest sounds. In patients with acute cough (“acute bronchitis”), a clinician could use these two factors as part of their overall clinical impression, which can be as accurate as a clinical decision rule in diagnosing acute respiratory infections [25]. An experienced clinician would use their overall clinical impression to judge if the patient is more likely to continue coughing, and could discuss with the patient the natural course and duration of a cough rather than immediately prescribing an antibiotic or corticosteroid.
A duration of cough greater than three weeks and dyspnea were both associated with an increased likelihood of receiving a CXR. The primary reasons to order a CXR are clinical suspicion of community-acquired pneumonia (CAP) or concern over malignancy, so this seems sensible. There was a trend for patients 60 years and older to be more likely to receive a CXR, which is also reasonable as the risk of both malignancy and CAP increase at this age.
Previous studies have shown that patients with acute LRTI but normal vital signs and a normal chest exam have a very low likelihood of CAP [26]. In our population, 58 patients had normal vital signs and a normal chest exam. While none had pneumonia, 14 received a chest radiograph, a rate only slightly lower than that of patients with abnormal vitals or lung exam.
We found a very high rate of systemic corticosteroid use, with 70% of all patients with a cough for one week or more receiving one. It is important to note that we excluded patients with asthma or COPD, and that of the 97 patients that had no wheezing on examination, 65 (67%) still received a systemic corticosteroid. A recent randomized trial found no benefit of corticosteroids for patients with uncomplicated acute bronchitis [13]. Current evidence supports corticosteroids to be limited to patients with wheezing from acute exacerbations of asthma [27] and selected patients with CAP [28]. This is therefore an important topic for further research to see if similarly high and inappropriate rates are seen in other urgent care centers, or in other contemporary ambulatory settings. It is also an important topic for continuing education, as physicians may not be familiar with recent negative studies on the topic.
Interestingly, increasing age was associated with receiving a cough suppressant in this population. Patients that were given a cough suppressant were a mean of 7.7 years older than those that did not. We were unable to find any literature addressing predictors of cough suppressant use in patients with prolonged cough. In general, there is insufficient evidence regarding the effectiveness of antitussive medication for the treatment of acute cough, and more generally of how acute respiratory tract infections (ARTI) are managed in the urgent care setting [14,15].
This study provides evidence that despite continued efforts to reduce antibiotic use in uncomplicated lower respiratory infections, they continue to be overprescribed. Approximately 95% of our population received an antibiotic, despite only four patients having CXR confirmed pneumonia. In addition, 70% received a systemic corticosteroid, which appears to represent a new trend in treatment of ARTI in the ambulatory setting. The implications of this are important, as a recent, population-based cohort study found an increased risk of adverse events from short courses of oral corticosteroids, including sepsis, venous thromboembolism and fractures [29].
The mean duration of cough (22 days) observed in this population of patients presenting with 7 or more days of cough was similar or slightly higher than that found in other studies [30,31]. Even though our inclusion criteria of 7 or more days of cough likely biased the duration to be longer, it was still within the range of 15 to 28 days found in a systematic review and well above the number of days patients expect their cough to resolve [5].
Given the recent increase in patients using emergency and urgent care over a primary care practice [16], we encourage additional observational prospective studies to determine if our findings are trends occurring at other urgent care settings. We also advocate for additional education to nromote the annronriate use of CXRs antibiotics corticosteroids, and cough suppressants and discussion with patients about the expectations for duration of illness.
Acknowledgements
The authors would like to recognize the medical director and staff of the Piedmont Healthcare Urgent Care Centers for their assistance in conducting this study.
Support and role of funding source
Funding provided by Institute Evidence-Based Health Professions Education. The funders provided compensation to patients for participation in the study and study materials including stock paper for diaries and data forms, and specimen collection swabs. The funding source had no role in the preparation or writing of the manuscript or any decisions on if or where to submit for publication.
Footnotes
Declarations of interest
None.
References
- [1].CDC. National Ambulatory Medical Care Survey: 2015 State and National Summary Tables CDC/National Center for Health Statistics; 2015. [Google Scholar]
- [2].Soler JK, Okkes I, Oskam S, Van Boven K, Zivotic P, Jevtic M, et al. The interpretation of the reasons for encounter ‘cough’ and ‘sadness’ in four international family medicine populations. Inform Prim Care 2012;20:25–39. [DOI] [PubMed] [Google Scholar]
- [3].Okkes IM, Oskam SK, Lamberts H. The probability of specific diagnoses for patients presenting with common symptoms to Dutch family physicians. J Fam Pract 2002; 51:31–6. [PubMed] [Google Scholar]
- [4].van Vugt SF, Verheij TJ, de Jong PA, Butler CC, Hood K, Coenen S, et al. Diagnosing pneumonia in patients with acute cough: clinical judgment compared to chest radiography. Eur Respir J 2013;42:1076–82. [DOI] [PubMed] [Google Scholar]
- [5].Ebell MH, Lundgren J, Youngpairoj S. How long does a cough last? Comparing patients’ expectations with data from a systematic review of the literature. Ann Fam Med 2013;11:5–13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [6].Ebell MH, Radke T. Antibiotic use for viral acute respiratory tract infections remains common. Am J Manag Care 2015;21:e567–75. [PubMed] [Google Scholar]
- [7].Suda KJ, Hicks LA, Roberts RM, Hunkler RJ, Danziger LH. A national evaluation of antibiotic expenditures by healthcare setting in the United States, 2009. J Antimicrob Chemother 2013;68:715–8. [DOI] [PubMed] [Google Scholar]
- [8].Fleming-Dutra KE, Hersh AL, Shapiro DJ, Bartoces M, Enns EA, File TM Jr, et al. Prevalence of in appropriate antibiotic prescriptions among US ambulatory care visits, 2010–2011. JAMA 2016;315:1864–73. [DOI] [PubMed] [Google Scholar]
- [9].Barnett ML, Linder JA. Antibiotic prescribing for adults with acute bronchitis in the United States, 1996–2010.JAMA 2014;311:2020–2. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [10].Lee GC, Reveles KR, Attridge RT, Lawson KA, Mansi IA, Lewis 2ndJS, et al. Outpatient antibiotic prescribing in the United States: 2000 to 2010. BMC Med 2014;12:96. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [11].Centre for Clinical Practice at N. National Institute for Health and Clinical Excellence: Guidance Respiratory tract infections-antibiotic prescribing: prescribing of antibiotics for self-limiting respiratory tract infections in adults and children in primary care. London: National Institute for Health and Clinical Excellence; 2008. [PubMed] [Google Scholar]
- [12].Irwin RS, Baumann MH, Bolser DC, Boulet LP, Braman SS, Brightling CE, et al. Diagnosis and management of cough executive summary: ACCP evidence-based clinical practice guidelines. Chest 2006;129:1s–23s. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [13].Hay AD, Little P, Harnden A, Thompson M, Wang K, Kendrick D, et al. Effect of oral prednisolone on symptom duration and severity in nonasthmatic adults with acute lower respiratory tract infection: a randomized clinical trial. JAMA 2017;318: 721–30. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [14].Holzinger F, Beck S, Dini L, Stoter C, Heintze C. The diagnosis and treatment of acute cough in adults. Dtsch Arztebl Int 2014;111:356–63. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [15].Schroeder K, Fahey T. Over-the-counter medications for acute cough in children and adults in ambulatory settings. Cochrane Database Syst Rev 2004(4). 10.1002/14651858.CD001831.pub2 Cd001831. [DOI] [PubMed] [Google Scholar]
- [16].Coster JE, Turner JK, Bradbury D, Cantrell A. Why do people choose emergency and urgent care services? A rapid review utilizing a systematic literature search and narrative synthesis. Acad Emerg Med 2017;24:1137–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [17].Coenen S, Michiels B, Renard D, Denekens J, Van Royen P. Antibiotic prescribing for acute cough: the effect of perceived patient demand. Br J Gen Pract 2006;56:183–90. [PMC free article] [PubMed] [Google Scholar]
- [18].Sirota M, Round T, Samaranayaka S, Kostopoulou O. Expectations for antibiotics increase their prescribing: causal evidence about localized impact. Health Psychol 2017;36:402–9. [DOI] [PubMed] [Google Scholar]
- [19].Welschen I, Kuyvenhoven M, Hoes A, Verheij T. Antibiotics for acute respiratory tract symptoms: patients’ expectations, GPs’ management and patient satisfaction. Fam Pract 2004;21:234–7. [DOI] [PubMed] [Google Scholar]
- [20].Macfarlane J, Holmes W, Macfarlane R, Britten N. Influence of patients’ expectations on antibiotic management of acute lower respiratory tract illness in general practice: questionnaire study. BMJ 1997;315:1211–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [[21].Coenen S, Francis N, Kelly M, Hood K, Nuttall J, Little P, et al. Are patientviews about antibiotics related to clinician perceptions, management and outcome? A multicountry study in outpatients with acute cough. PLoS One 2013;8:e76691. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [22].Burnham K Model selection and multimodel inference: a practical information-theoretic approach. 2nd ed. Springer-Verlag; New York; 2002. [Google Scholar]
- [23].Burnham KP, Anderson DR. Multimodel inference understanding AIC and BIC in model selection. Sociol Methods Res 2004;33:261–304. [Google Scholar]
- [24].Symonds MRE, Moussalli A. A brief guide to model selection, multimodel inference and model averaging in behavioural ecology using Akaike’s information criterion. Behav Ecol Sociobiol 2011;65:13–21. [Google Scholar]
- [25].Dale AP, Marchello C, Ebell M. The accuracy of the overall clinical impression for the diagnosis of pharyngitis, sinusitis, and pneumonia: a meta-analysis; 2018. [Unpublished. Currently in peer review]. [Google Scholar]
- [26].Marchello C, Ebell M, Dale AP, Harvill E, Shen Y, Whalen CC. Signs and symptoms that rule out community-acquired pneumonia in outpatient adults: a systematic review and meta-analysis. J Am Board Fam Med 2019. [Accepted; in press]. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [27].Rowe BH, Spooner C, Ducharme FM, Bretzlaff JA, Bota GW. Early emergency department treatment of acute asthma with systemic corticosteroids. Cochrane Database Syst Rev 2001(1). 10.1002/14651858.CD002178 Cd002178. [DOI] [PubMed] [Google Scholar]
- [28].Stern A, Skalsky K, Avni T, Carrara E, Leibovici L, Paul M. Corticosteroids for pneumonia. Cochrane Database Syst Rev 2017;12:Cd007720. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [29].Waljee AK, Rogers MA, Lin P, Singal AG, Stein JD, Marks RM, et al. Short term use of oral corticosteroids and related harms among adults in the United States: population based cohort study. BMJ 2017;357:j1415. [DOI] [PMC free article] [PubMed] [Google Scholar]
- [30].Butler CC, Hood K, Kelly MJ, Goossens H, Verheij T, Little P, et al. Treatment of acute cough/lower respiratory tract infection by antibiotic class and associated outcomes: a 13 European country observational study in primary care.J Antimicrob Chemother 2010;65:2472–8. [DOI] [PubMed] [Google Scholar]
- [31].Little P, Rumsby K, Kelly J, Watson L, Moore M, Warner G, et al. Information leaflet and antibiotic prescribing strategies for acute lower respiratory tract infection: a randomized controlled trial. JAMA 2005;293:3029–35. [DOI] [PubMed] [Google Scholar]