An interventional study of PRACTICE-LRTI multimodal antimicrobial stewardship programme in Bronchitis, Acute, Community-acquired Pneumonia and Respiratory Tract Infections, sponsored by Tampere University. Not yet recruiting at 1 site in Finland. Open to participants aged Up to 17 Years. Per ClinicalTrials.gov, last updated 2026-10-26.
Sponsored by Tampere University · Not applicable, Interventional, and Health services research
Children with lower respiratory tract infection (LRTI) are still often prescribed antibiotics in ways that depart from national guidelines. Acute viral bronchitis is treated with antibiotics although none are needed, and community-acquired pneumonia (CAP) is often treated with a second-choice antibiotic, most often a macrolide, instead of first-line amoxicillin. In the Terveystalo network in Finland, about one third of bronchitis visits and one fifth of CAP visits in 2014 to 2020 led to an antibiotic prescription, and macrolides were the most commonly prescribed agent for bronchitis.
PRACTICE-LRTI is a nationwide quality-improvement and implementation study in which all clinics of the network receive a multimodal antimicrobial stewardship programme aimed at prescribing clinicians (paediatricians, general practitioners, ear, nose and throat specialists and other physicians). The programme combines (1) a concise local version of the Finnish Current Care Guideline for paediatric LRTI; (2) education for physicians and caregivers (printed and intranet materials, videos, webinars, unit-level meetings, online courses and commitment posters); (3) light-touch electronic health record reminders and a one-click supportive-care order set; and (4) a real-time prescribing dashboard with a structured feedback cascade that escalates over three years from network-wide progress reports to targeted outreach to non-improving units and, finally, personalised letters and phone calls to persistently non-concordant prescribers. A one-page guide prompts clinicians to question unverified penicillin-allergy labels.
The programme is delivered to the whole network at the same time, so no clinics are randomised to a control condition. Instead, prescribing is followed as a monthly time series from 2014 onward, and the change after each of three prespecified intervention steps is estimated by segmented regression against the projected pre-intervention trend. The co-primary outcomes are the proportion of acute bronchitis episodes in which a systemic antibiotic is prescribed and the proportion of antibiotic-treated outpatient CAP episodes in which a non-first-line agent is prescribed. The co-primary outcomes are assessed at Month 36 after launch; their Month 12 and Month 48 assessments are secondary outcomes. Other secondary outcomes include first-line amoxicillin and macrolide use, 5-day treatment duration, variation between units and physicians, diagnostic testing, first-line prescribing in children with a penicillin-allergy label, and intervention cost. Safety is monitored through 14-day respiratory revisits, hospital referrals and hospitalisations. The active intervention lasts 3 years and is followed by 1 year of follow-up to assess durability.
This is a quality-improvement and implementation study embedded in routine care, not a trial of a treatment. The intervention is directed at physicians; no child is recruited, contacted or examined outside routine care, and all outcome data come from the network's electronic health record.
Design. Prospective, pragmatic, multicentre quality-improvement and implementation study with a multi-step interrupted time-series (ITS) evaluation and continuous electronic health record (EHR) surveillance in a nationwide private healthcare network in Finland (about 300 clinics, about 300,000 paediatric outpatient visits per year). Cluster randomisation is not feasible because the network has centralised medical leadership, a unified EHR, shared quality surveillance and a common intranet, so company-wide materials would contaminate any control clusters; for the same reason the managerial components (unit feedback, outreach, letters and calls) cannot be staggered or randomised across units. The counterfactual is provided by an extended pre-intervention baseline (published 2014 to 2020 series plus a monthly series from January 2021 to December 2026, the month before Day 0) and segmented-regression modelling of level and trend.
Intervention. Four components delivered in three annual graduated steps. Step 1 (Months 1 to 12): company-wide launch of the local guideline, workflow aids, physician and caregiver videos, nationwide online meetings, in-person unit meetings, online courses, caregiver leaflet, EHR reminders, live dashboard, quarterly progress reports and a first round of direct feedback from the Head of Pediatrics and unit Chief Medical Officers. Step 2 (Months 13 to 24): targeted outreach to non-improving units and physicians, repeat unit meetings, intensified direct feedback. Step 3 (Months 25 to 36): personalised letters to persistently non-concordant prescribers, phone calls where letters do not change practice, recognition of high performers. Year 4 (Months 37 to 48): follow-up without new active push. Delivery of every component is logged prospectively and reported as process and fidelity measures alongside the outcome measures: online-course completion, webinar attendance and unit-meeting delivery (education reach), materials deployment, EHR reminder display and supportive-care order-set use, dashboard access, and the number and timing of direct feedback sessions, personalised letters and phone calls with the proportion of planned leadership acts delivered on schedule. From Month -6 every leadership act (report, meeting, letter, call) is recorded with timestamp, deliverer, recipient and planned date in a leadership log.
Unit of analysis. The LRTI episode: the index visit plus related visits within 14 days for the same child and diagnostic group. Children are analysed in three age groups (under 5, 5 to 9 and 10 to 17 years) and stratified by sex, physician specialty and unit or region.
Analysis. Primary analysis: segmented binomial regression of monthly proportions with a level-change indicator and trend-change term at Day 0 and prespecified secondary change points at Month 13, Month 25 and Month 37, harmonic seasonal terms, COVID-period indicators, and correction for autocorrelation (Newey-West or Prais-Winsten; ARIMA as robustness check). Primary estimand: absolute difference between observed and counterfactual proportions at Month 36, with 95% confidence intervals (delta method, verified by parametric bootstrap); the same estimand at Month 12 and Month 48 is secondary. The two co-primary outcomes are tested conjunctively at two-sided alpha 0.05 each. Secondary analyses: three-level multilevel logistic regression (episodes within physicians within units) with variance components summarised as intraclass correlation and median odds ratio; funnel plots and statistical process control charts; a prespecified physician-level dose-response analysis of educational exposure and a unit-level engagement-response analysis; matched before-after comparison of personalised-feedback recipients with negative-control selection in a baseline year; an explanatory leadership-effect analysis (within-clinician comparison across units for clinicians who work in more than one unit, leadership dose-response, and an event-study of the dates on which unit feedback was delivered); prespecified subgroup, sensitivity and quantitative bias analyses, including quarterly aggregation of the CAP series, an indicator term for the preparatory-phase months (July to December 2026), a placebo interruption date and a negative-control outcome (antibiotic prescribing for paediatric urinary tract infection). Balancing and safety outcomes (14-day respiratory revisit, hospital referral, 30-day LRTI hospitalisation, CRP testing or chest radiography in bronchitis episodes as a substitution signal, antibiotic-allergy labels, and visit-level patient satisfaction) are followed on monthly run charts with prespecified signal thresholds and analysed with the same ITS models. A prespecified health-economic evaluation (cost-consequence analysis with cost-effectiveness ratios, provider perspective, CHEERS 2022) quantifies intervention costs by component and step against antibiotic-related savings derived from the ITS counterfactual.
Sample size. All eligible episodes are included (about 10,000 bronchitis and 2,000 to 2,500 CAP episodes per year, of which about 450 to 560 CAP episodes are antibiotic-treated and form the denominator of the CAP co-primary outcome), giving precision to detect absolute changes of about 3 percentage points in bronchitis prescribing and about 5 points in CAP agent selection. No formal power calculation governs inclusion.
Registration and reporting. The statistical analysis plan (version 1.0) was frozen on 8 October 2026, before Day 0, and is attached to this record as a study document; the protocol is submitted as a protocol paper. Analyses are conducted by the principal investigator with Terveystalo's statistical team and the research group's statistician; the analysis plan, analysis code and co-intervention log are published with the results. Reporting follows SQUIRE 2.0, TIDieR and EPOC guidance for interrupted time series.
273 studies on the registry are indexed under Bronchitis; 27 are open to participants now.
This study's planned enrollment of 50,000 is above the median of 120 across 188 interventional studies indexed under Bronchitis.
Browse Bronchitis studies →Tampere University is the lead sponsor of 63 studies on the registry; 11 are open to participants now.
Counted across the registry records on this site, refreshed daily.
Patient-episode inclusion criteria:
Patient-episode exclusion criteria:
Clinicians: all physicians who record an eligible LRTI visit in the network during the study period receive the programme; there are no clinician exclusion criteria. The intervention targets prescribing clinicians; outcomes are measured on patient episodes identified from routinely collected electronic health record data without recruitment or contact.
All prescribing clinicians and all clinics of the Terveystalo network receive the four-component programme in three graduated annual steps. Outcomes are measured on all eligible paediatric LRTI episodes recorded in the network's EHR before, during and after the intervention.
Behavioral: PRACTICE-LRTI multimodal antimicrobial stewardship programme
Four components in three annual steps. (1) Guideline reinforcement: concise local version of the Finnish Current Care Guideline for paediatric LRTI (no antibiotics for bronchitis; amoxicillin 40 to 45 mg/kg/day for 5 days as first-line for uncomplicated CAP; macrolide only for suspected Mycoplasma; indications for CRP and chest radiography). (2) Multimodal education: printed and intranet materials, physician and caregiver videos, webinars, unit meetings, online courses, commitment posters, caregiver leaflet. (3) EHR support: non-blocking diagnosis-specific reminders, allergy-label prompt, one-click supportive-care order set. (4) Real-time dashboard and feedback cascade: physician-, unit- and network-level indicators on SPC charts; quarterly reports; direct feedback from the Head of Pediatrics and unit Chief Medical Officers; in Step 3, personalised letters and calls to persistently non-concordant prescribers. A one-page guide prompts questioning of unverified penicillin-allergy labels.
Also known as: Guideline reinforcement, Multimodal clinician and caregiver education, EHR reminders and supportive-care order sentence, Real-time prescribing dashboard with audit and feedback
Proportion of acute bronchitis episodes with a systemic antibiotic prescription
Proportion of eligible acute bronchitis episodes (index visit plus related visits within 14 days for the same child) with at least one systemic antibiotic prescription, measured in Month 36 after programme launch (end of Step 3). Unit of measure: percentage of episodes. The intervention effect is reported as the absolute difference in percentage points, with 95% confidence interval, between the observed proportion and the counterfactual proportion projected from the pre-launch monthly series (published 2014 to 2020 series extended month by month to the month before launch), estimated by segmented binomial regression. Target: below 15% (minimum acceptable below 20%) from a baseline of 32%.
Time frame: Month 36 after programme launch
Proportion of antibiotic-treated outpatient community-acquired pneumonia episodes with a non-first-line antibiotic
Proportion of antibiotic-treated outpatient community-acquired pneumonia (CAP) episodes in which a non-first-line agent is prescribed (a macrolide, or a broad-spectrum agent without a documented indication; amoxicillin is first-line), measured in Month 36 after programme launch (end of Step 3). Unit of measure: percentage of antibiotic-treated CAP episodes. The intervention effect is reported as the absolute difference in percentage points, with 95% confidence interval, between the observed proportion and the counterfactual proportion projected from the pre-launch monthly series, estimated by segmented binomial regression. Corresponds to the target of increasing first-line amoxicillin from 65% to above 80% of antibiotic-treated CAP.
Time frame: Month 36 after programme launch
Proportion of acute bronchitis episodes with a systemic antibiotic prescription at Month 12
Same measurement and analysis as the first co-primary outcome, assessed in Month 12 after programme launch (end of Step 1). Unit of measure: percentage of episodes; effect reported as the absolute difference in percentage points from the counterfactual, with 95% confidence interval.
Time frame: Month 12 after programme launch
Proportion of acute bronchitis episodes with a systemic antibiotic prescription at Month 48
Same measurement and analysis as the first co-primary outcome, assessed in Month 48 after programme launch (end of the 1-year follow-up without new active intervention), as the measure of durability. Unit of measure: percentage of episodes; effect reported as the absolute difference in percentage points from the counterfactual, with 95% confidence interval.
Time frame: Month 48 after programme launch
Proportion of antibiotic-treated outpatient community-acquired pneumonia episodes with a non-first-line antibiotic at Month 12
Same measurement and analysis as the second co-primary outcome, assessed in Month 12 after programme launch (end of Step 1). Unit of measure: percentage of antibiotic-treated CAP episodes; effect reported as the absolute difference in percentage points from the counterfactual, with 95% confidence interval.
Time frame: Month 12 after programme launch
Proportion of antibiotic-treated outpatient community-acquired pneumonia episodes with a non-first-line antibiotic at Month 48
Same measurement and analysis as the second co-primary outcome, assessed in Month 48 after programme launch (end of the 1-year follow-up), as the measure of durability. Unit of measure: percentage of antibiotic-treated CAP episodes; effect reported as the absolute difference in percentage points from the counterfactual, with 95% confidence interval.
Time frame: Month 48 after programme launch
Proportion of antibiotic-treated outpatient community-acquired pneumonia episodes treated with amoxicillin
Proportion of antibiotic-treated outpatient CAP episodes in which amoxicillin (ATC J01CA04) is the prescribed antibiotic, measured in Month 36 after programme launch. Unit of measure: percentage of antibiotic-treated CAP episodes. Analysed by segmented regression against the pre-launch counterfactual. Target: above 80%.
Time frame: Month 36 after programme launch
Proportion of antibiotic-treated outpatient community-acquired pneumonia episodes with a macrolide prescription
Proportion of antibiotic-treated outpatient CAP episodes in which a macrolide (ATC J01FA) is prescribed, measured in Month 36 after programme launch. Unit of measure: percentage of antibiotic-treated CAP episodes. Analysed by segmented regression against the pre-launch counterfactual. Target: below 10%.
Time frame: Month 36 after programme launch
Proportion of acute bronchitis episodes with a macrolide prescription
Proportion of eligible acute bronchitis episodes in which a macrolide (ATC J01FA) is prescribed, measured in Month 36 after programme launch. Unit of measure: percentage of episodes. Analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of community-acquired pneumonia amoxicillin courses with a prescribed duration of 5 days or less
Proportion of amoxicillin courses prescribed for outpatient CAP with a prescribed duration of 5 days or less (guideline-concordant duration), measured in Month 36 after programme launch. Unit of measure: percentage of CAP amoxicillin courses. Analysed by segmented regression against the pre-launch counterfactual. Target: above 50%.
Time frame: Month 36 after programme launch
Median odds ratio between units for antibiotic prescribing in acute bronchitis
Median odds ratio (MOR) for the unit level from a three-level multilevel logistic regression (eligible acute bronchitis episodes within physicians within units) of systemic antibiotic prescribing, estimated for the Step 3 period. The same model is fitted to the pre-launch baseline period and to each study period so that the change in between-unit variation can be described. Unit of measure: odds ratio (dimensionless).
Time frame: Step 3 period, Months 25 to 36 after programme launch
Median odds ratio between physicians for antibiotic prescribing in acute bronchitis
Median odds ratio (MOR) for the physician level from the same three-level multilevel logistic regression of systemic antibiotic prescribing in eligible acute bronchitis episodes, estimated for the Step 3 period and, for comparison, for the pre-launch baseline period and each study period. Unit of measure: odds ratio (dimensionless).
Time frame: Step 3 period, Months 25 to 36 after programme launch
Median odds ratio between units for non-first-line antibiotic prescribing in community-acquired pneumonia
Median odds ratio (MOR) for the unit level from a three-level multilevel logistic regression (antibiotic-treated outpatient CAP episodes within physicians within units) of non-first-line agent prescribing, estimated for the Step 3 period and, for comparison, for the pre-launch baseline period and each study period. Unit of measure: odds ratio (dimensionless).
Time frame: Step 3 period, Months 25 to 36 after programme launch
Median odds ratio between physicians for non-first-line antibiotic prescribing in community-acquired pneumonia
Median odds ratio (MOR) for the physician level from the same three-level multilevel logistic regression of non-first-line agent prescribing in antibiotic-treated outpatient CAP episodes, estimated for the Step 3 period and, for comparison, for the pre-launch baseline period and each study period. Unit of measure: odds ratio (dimensionless).
Time frame: Step 3 period, Months 25 to 36 after programme launch
Proportion of lower respiratory tract infection episodes coded as pneumonia or unspecified lower respiratory tract infection
Proportion of all eligible lower respiratory tract infection (LRTI) episodes coded as community-acquired pneumonia or unspecified LRTI rather than acute bronchitis, measured in Month 36 after programme launch, to detect shifts in diagnostic labelling during the intervention. Unit of measure: percentage of eligible LRTI episodes. Analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of lower respiratory tract infection episodes with diagnostic testing
Proportion of eligible LRTI episodes in which any of C-reactive protein (CRP) testing, rapid viral or influenza testing, or chest radiography is ordered at the index visit, each episode counted once regardless of the number or type of tests, measured in Month 36 after programme launch. Unit of measure: percentage of eligible LRTI episodes. Analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of first-line antibiotic treatment among lower respiratory tract infection episodes with a documented penicillin-class allergy label
Among eligible LRTI episodes in which the child has a documented penicillin-class allergy label on file at the visit, the proportion treated with amoxicillin or another first-line agent, measured in Month 36 after programme launch. Unit of measure: percentage of allergy-labelled LRTI episodes. Analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Total intervention cost
Prospectively logged cost of delivering the programme (micro-costing of personnel time, materials, EHR development and feedback activities), summed across components and steps from the start of the preparatory phase to the end of the follow-up year. Unit of measure: euros. Reported with a breakdown by component and step (provider perspective, CHEERS 2022).
Time frame: Cumulative from Month -6 (start of the preparatory phase) through Month 48 after programme launch
Net programme cost
Total intervention cost minus the direct antibiotic cost savings attributable to the programme, where savings are derived from the difference between observed and counterfactual antibiotic prescribing in the interrupted time series. Unit of measure: euros, with bootstrap 95% confidence interval (provider perspective).
Time frame: Cumulative from Month -6 (start of the preparatory phase) through Month 48 after programme launch
Cost per averted acute bronchitis antibiotic course
Net programme cost divided by the number of systemic antibiotic courses for acute bronchitis averted relative to the interrupted time series counterfactual over the same period (cost-consequence analysis, provider perspective). Unit of measure: euros per averted antibiotic course, with bootstrap 95% confidence interval.
Time frame: Cumulative from Month -6 (start of the preparatory phase) through Month 48 after programme launch
Cost per additional first-line-treated community-acquired pneumonia episode
Net programme cost divided by the number of additional antibiotic-treated outpatient CAP episodes treated with a first-line agent relative to the interrupted time series counterfactual over the same period (cost-consequence analysis, provider perspective). Unit of measure: euros per additional first-line-treated episode, with bootstrap 95% confidence interval.
Time frame: Cumulative from Month -6 (start of the preparatory phase) through Month 48 after programme launch
Proportion of lower respiratory tract infection episodes with a respiratory revisit within 14 days (balancing outcome)
Proportion of eligible LRTI episodes followed by a respiratory revisit within 14 days of the index visit, measured in Month 36 after programme launch. Unit of measure: percentage of eligible LRTI episodes. Followed monthly on run charts from launch and analysed by segmented regression against the pre-launch counterfactual. Prespecified safety signal: intervention-attributable absolute increase exceeding 1 percentage point.
Time frame: Month 36 after programme launch
Proportion of lower respiratory tract infection episodes with a hospital referral (balancing outcome)
Proportion of eligible LRTI episodes with a hospital referral at or after the index visit within the episode, measured in Month 36 after programme launch. Unit of measure: percentage of eligible LRTI episodes. Followed monthly on run charts from launch and analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of lower respiratory tract infection episodes with hospitalisation for lower respiratory tract infection within 30 days (balancing outcome)
Proportion of eligible LRTI episodes followed by hospitalisation for LRTI within 30 days of the index visit, measured in Month 36 after programme launch. Unit of measure: percentage of eligible LRTI episodes. Followed monthly on run charts from launch and analysed by segmented regression against the pre-launch counterfactual. Any statistically significant intervention-attributable increase is a prespecified safety signal.
Time frame: Month 36 after programme launch
Proportion of acute bronchitis episodes with C-reactive protein testing or chest radiography (substitution signal)
Proportion of eligible acute bronchitis episodes in which CRP testing or chest radiography, or both, is ordered at the index visit, each episode counted once, measured in Month 36 after programme launch. Unit of measure: percentage of acute bronchitis episodes. An increase while antibiotic prescribing falls is interpreted as testing substituting for antibiotics rather than guideline-concordant care. Analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of lower respiratory tract infection episodes with a documented penicillin-class allergy label (balancing outcome)
Proportion of eligible LRTI episodes in which the child has a documented penicillin-class allergy label on file at the visit, measured in Month 36 after programme launch. Unit of measure: percentage of eligible LRTI episodes. Followed monthly from launch and analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Mean Net Promoter Score for lower respiratory tract infection visits (balancing outcome)
Mean of the Terveystalo visit-level Net Promoter Score question (0 to 10 scale, higher is more favourable) among eligible LRTI visits with a returned score, measured in Month 36 after programme launch. Unit of measure: score on a 0 to 10 scale. Followed monthly from launch and analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Proportion of promoters among lower respiratory tract infection visits (balancing outcome)
Proportion of eligible LRTI visits with a returned Net Promoter Score of 9 or 10 (promoters), measured in Month 36 after programme launch. Unit of measure: percentage of LRTI visits with a returned score. Followed monthly from launch and analysed by segmented regression against the pre-launch counterfactual.
Time frame: Month 36 after programme launch
Within-clinician difference in acute bronchitis antibiotic prescribing between units with and without prior unit-level leadership exposure (explanatory)
Among clinicians with eligible acute bronchitis episodes in two or more units in the same calendar month, the difference in the proportion of episodes with a systemic antibiotic prescription between units in which unit-level leadership exposure (unit Chief Medical Officer feedback delivered or unit meeting held, from the prospective leadership log) has already occurred and units in which it has not, pooled over Months 1 to 36 with clinician fixed effects. Unit of measure: percentage points. Reported as an explanatory lower bound of the site-bound leadership effect.
Time frame: From programme launch through Month 36 after programme launch
Within-clinician difference in non-first-line antibiotic prescribing in community-acquired pneumonia between units with and without prior unit-level leadership exposure (explanatory)
Among clinicians with antibiotic-treated outpatient CAP episodes in two or more units in the same calendar month, the difference in the proportion of episodes with a non-first-line agent between units with and without prior unit-level leadership exposure (defined as above), pooled over Months 1 to 36 with clinician fixed effects. Unit of measure: percentage points. Reported as an explanatory lower bound of the site-bound leadership effect.
Time frame: From programme launch through Month 36 after programme launch
Plan to share: No — No data are available at the protocol stage. Episode-level data are held by Terveystalo and cannot be shared by the investigators because they are confidential health data; third-party access to the underlying data is governed by Terveystalo and, where applicable, requires a Findata permit under the Finnish secondary-use act. The SAP, code lists, indicator definitions and intervention materials are provided as online supplemental material; analysis code will be provided at publication of results.
Start is estimated for Jan 2027; primary completion for Dec 2029, 35 months later.
Of 33 interventional studies with no phase under Bronchitis, half reached primary completion within 29 months of starting (the median), counting those still running; a quarter within 9 and three quarters within 38. This study's planned 35 months is above that.
How these figures are worked out, and which studies are left out of them.
From the registry record's own update history. This site started tracking changes on Sep 25, 2026; for anything earlier, see the record history on ClinicalTrials.gov ↗
This study is not yet recruiting, as verified in Oct 2026. You cannot join it, but the record below documents what was studied.
Get an email when the registry record changes — status, dates, results — or when someone posts here.
Sign in to followQuestions and observations about this study, from anyone following it. Not medical advice, and not a channel to the study team — their contact details are on the registry record.
Sign in to join the discussion. Reading takes no account; posting does. You choose a display name, and a pseudonym is the default.
Nothing here yet. If you are running this trial, taking part in it, or weighing whether to, this is the place to say so.
Tampere University