key: cord-0871386-u3jlmbct
authors: Lhopitallier, Loïc; Kronenberg, Andreas; Meuwly, Jean-Yves; Locatelli, Isabella; Mueller, Yolanda; Senn, Nicolas; D’Acremont, Valérie; Boillat-Blanco, Noémie
title: Procalcitonin and lung ultrasonography point-of-care testing to determine antibiotic prescription in patients with lower respiratory tract infection in primary care: pragmatic cluster randomised trial
date: 2021-09-21
journal: BMJ
DOI: 10.1136/bmj.n2132
sha: 50d51b1fac1b8e8196169ec6088852c8e21e5397
doc_id: 871386
cord_uid: u3jlmbct

OBJECTIVE: To assess whether point-of care procalcitonin and lung ultrasonography can safely reduce unnecessary antibiotic treatment in patients with lower respiratory tract infections in primary care. DESIGN: Three group, pragmatic cluster randomised controlled trial from September 2018 to March 2020. SETTING: 60 Swiss general practices. PARTICIPANTS: One general practitioner per practice was included. General practitioners screen all patients with acute cough; patients with clinical pneumonia were included. INTERVENTIONS: Randomisation in a 1:1:1 of general practitioners to either antibiotics guided by sequential procalcitonin and lung ultrasonography point-of-care tests (UltraPro; n=152), procalcitonin guided antibiotics (n=195), or usual care (n=122). MAIN OUTCOMES: Primary outcome was proportion of patients in each group prescribed an antibiotic by day 28. Secondary outcomes included duration of restricted activities due to lower respiratory tract infection within 14 days. RESULTS: 60 general practitioners included 469 patients (median age 53 years (interquartile range 38-66); 278 (59%) were female). Probability of antibiotic prescription at day 28 was lower in the procalcitonin group than in the usual care group (0.40 v 0.70, cluster corrected difference −0.26 (95% confidence interval −0.41 to −0.10)). No significant difference was seen between UltraPro and procalcitonin groups (0.41 v 0.40, −0.03 (−0.17 to 0.12)). The median number of days with restricted activities by day 14 was 4 days in the procalcitonin group and 3 days in the usual care group (difference 1 day (95% confidence interval −0.23 to 2.32); hazard ratio 0.75 (95% confidence interval 0.58 to 0.97)), which did not prove non-inferiority. CONCLUSIONS: Compared with usual care, point-of-care procalcitonin led to a 26% absolute reduction in the probability of 28 day antibiotic prescription without affecting patients’ safety. Point-of-care lung ultrasonography did not further reduce antibiotic prescription, although a potential added value cannot be excluded, owing to the wide confidence intervals. TRIAL REGISTRATION: ClinicalTrials.gov NCT03191071.

: Full list of inclusion and exclusion criteria for the recruitment of participants in the UltraPro study.

All general practitioners participated in a 2-hour seminar on the rationale (antibiotic resistance, epidemiology of pneumonia in Switzerland) of the study and its procedures.

General practitioners in the UltraPro and procalcitonin groups participated in an additional 2-hour training session. Topics included management of community-acquired pneumonia in primary care and the use of procalcitonin and lung ultrasound to guide antibiotic prescription.

General practitioners in the UltraPro group participated in an additional lung ultrasound half-day training session to achieve independent practice and appropriately identify lung consolidation. Topics included ultrasound physics, ultrasound equipment, probe positioning, image recording and interpretation using a phantom simulator (CAE Healthcare©).

Before study start, the study team visited all general practitioners at their own practices and trained the medical assistants in study procedures and point-of-care procalcitonin measurement. Figure S1 : Description of the three arms of the study and of the UltraPro algorithm. PCT denotes procalcitonin. GPs denotes general practitioners.

Patients reported each day on six items (cough, phlegm, shortness of breath, sleep disturbance, impairment of normal daily activities and feeling unwell) on a Likert scale (1 to 6). By summing the values, we obtained a daily composite symptom score for each patient.

Participants reported either using a paper questionnaire or a web-based questionnaire. Participants filled the daily diary until symptom resolution, up to a maximum of 28 days.

Simulations used to determine the sample size are shown in the following figures: Figure S2 : Statistical simulations used to determine the sample size. Each line corresponds to a number of patients per general practitioner as indicated on the left-hand side. The figure on the left represents the simulation assuming a 15% difference between the proportion of patients prescribed an antibiotic for UltraPro (30%) compared to procalcitonin (45%). The figure on the right represents the simulation assuming a 15% difference between the proportion of patients prescribed an antibiotic for procalcitonin (45%) compared to usual care (60%). The intra-cluster coefficient estimate (0.06) used to account from clustering was derived from previously published data. 2 FP denotes family practitioner or general practitioner. In order to account for variability of cluster size within a group, we chose each cluster size from a positive distribution with defined mean and standard deviation. The variability of the cluster size had only a very small impact on the final sample size: given the number of clusters and the average cluster size, increasing variability parameter in the simulations did not generate lower powers.

We initially recruited 14 general practitioners per group and a mean of 15 patients per general practitioner in each group (210 patients per arm for a total of 630 patients). Due to lower than expected recruitment rates, the protocol was amended (v 6.0, 9 th of July 2019) and the number of general practitioners per group increased to 20.

The calculated sample size of 60 general practitioners with a mean of 10 patients per general practitioners (total 600 participants) further guarantees a power of 80% to prove non-inferiority in terms of duration of activities restriction by day 14 of UIltraPro group versus procalcitonin group, and of procalcitonin group versus usual care group (non-inferiority margin of 1 day, standard deviation of 4 days). We chose this standard deviation based on a previous publication. 1 For the sake of simplicity, this calculation assumes no intra-class correlation for this secondary outcome.

We calculated the effective power of the two comparisons using the number of clusters, the mean cluster size and the range of cluster sizes observed in each randomization group (UltraPro: n. of clusters =19, mean cluster size=8, range=1-20; procalcitonin: n. of clusters =19, mean cluster size=10, range=1-22; Usual care: n. of clusters =17, mean cluster size=7, range=1-15). We also used the observed ICC of the comparison between procalcitonin and Usual Care (ICC=0.15) and between UltraPro and procalcitonin (ICC=0.096), both larger than the assumed ICC=0.06. We found that the power for the first comparison (procalcitonin and usual care) was of 0.55, while the power for the second comparison (UltraPro and procalcitonin) was of 0.7. Figure S3 : Weekly rates of inclusion and screening over time. Due to an issue with the eCRF recording screening dates in the early weeks of the study, there were no proper date records of the first few weeks of screening.

The following tables shows the baseline characteristics of the patients with missing primary outcome data. We assumed the data to be missing at random. Table S2 : Baseline characteristics of participants according to whether or not they have missing data relating to the primary outcome of the study. COPD denotes chronic obstructive pulmonary disease, SD denotes standard deviation, IQR denotes inter-quartile range. Table S3 : Effect of procalcitonin versus usual care, and of UltraPro versus procalcitonin on antibiotic prescription at day 28 of follow-up, the analyses are performed on the per protocol population and on the intention to treat population with varying ways to deal with missing outcome data. Of note, three general practitioners were excluded from the per protocol analysis as all patients they included had missing outcome data.

95% CI denotes 95 percent confidence interval. ICC denotes intra-cluster coefficient. The estimated difference is corrected for cluster size. The following variables were used in the imputation model: sex, age, active smoking, alcohol abuse, heart failure, diabetes, chronic obstructive pulmonary disease, asthma, active malignancy, cough duration, sputum production, history of fever, duration of fever, history of dyspnoea, history of chest pain, blood pressure, heart rate, respiratory rate, temperature, SpO2 and abnormal auscultation. The imputation model did not take in account the cluster effect. The following variables were used for the adjusted multivariate mixed effect model: age, asthma, chronic obstructive pulmonary disease, sputum production, history of fever, history of dyspnoea, history of chest pain, CRB-65. Apart from the multivariate mixed effect model, all odds ratio reported are unadjusted. The variables for the adjusted and imputation model were chosen for being relevant and clinically sound baseline patient characteristics.

The p-value pertains to the odds ratio. Table S4 : Details of the class of antibiotics prescribed at day 0 by the general practitioners and the adhesion to the prescribed treatment by day 7. The regimens prescribed as other included one treatment with a combination of beta-lactam/macrolide and one treatment of clindamycin. IQR denotes inter-quartile range. Table S7 : Details of the various components of the composite score of clinical failure by day 7. This analysis is performed on the per protocol population. All hospitalisations recorded occurred within 7 days of enrolment, in the UltraPro group one patient was admitted due to acute appendicitis, one due to acute hepatitis and one due to tympanic perforation, in the Procalcitonin group one of the patients was admitted due to an acute psychotic episode. All other hospitalisations were attributed to a lower respiratory tract infection. Table S8 : Effect of procalcitonin versus usual care, and of UltraPro versus procalcitonin on serious adverse outcome (composite of death and hospitalisation at day 28). This analysis is performed on the per protocol population. ICC denotes intra-cluster coefficient. The estimated difference is corrected for cluster size.

The p-value pertains to the odds ratio. Figure S4 : Effect of the interventions on the prescription of chest X-ray at day 0 between the 3 study groups. There was no significant difference in the proportion of new lung infiltrates observed by the general practitioners on the chest X-ray between the usual care, procalcitonin and Table S9 : Prescriptions of C-Reactive protein and white blood count at day 0 according to study group. There is no statistical difference between the groups. Table S10 : Patient satisfaction regarding the consultation at day 0. Satisfaction was measured using sub-questions from the Visit-specific Satisfaction Instrument, previously validated for use in French. 4 by standardized phone interview at day 7. There were no statistically significant differences between groups for all four questions.

Procalcitonin-guided antibiotic use vs a standard approach for acute respiratory tract infections in primary care

Effect of point of care testing for C reactive protein and training in communication skills on antibiotic use in lower respiratory tract infections: cluster randomised trial

Assessment of oxygenation and comorbidities improves outcome prediction in patients with community-acquired pneumonia with a low CRB-65 score

Cultural adaptation and validation of questionnaires measuring satisfaction with the French health system