Temporarily quadrupling the dose of inhaled steroid to prevent asthma exacerbations: FAST

Article history

The research reported in this issue of the journal was funded by the HTA programme as project number 10/143/01. The contractual start date was in March 2013. The draft report began editorial review in September 2017 and was accepted for publication in July 2018. The authors have been wholly responsible for all data collection, analysis and interpretation, and for writing up their work. The HTA editors and publisher have tried to ensure the accuracy of the authors’ report and would like to thank the reviewers for their constructive comments on the draft document. However, they do not accept liability for damages or losses arising from material published in this report.

Declared competing interests of authors

David Price reports board membership fees paid to the Observational and Pragmatic Research Institute from Aerocrine AB, Amgen Inc., AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Chiesi Farmaceutici S.p.A., Mylan N.V., Mundipharma GmbH, Napp Pharmaceutical Group Ltd, Novartis Pharmaceutical Company and Teva Pharmaceutical Industries Ltd; consultancy agreement fees paid to the Observational and Pragmatic Research Institute from Almirall S.A., Amgen Inc., AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Chiesi Farmaceutici S.p.A., GlaxoSmithKline plc, Mylan N.V., Mundipharma GmbH, Napp Pharmaceutical Group Ltd, Novartis Pharmaceutical Company, Pfizer Inc., Teva Pharmaceutical Industries and Theravance Biopharma; grants from Aerocrine AB, AKL Research and Development Ltd, AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, the British Lung Foundation, Chiesi Farmaceutici S.p.A., Mylan N.V., Mundipharma GmbH, Napp Pharmaceutical Group Ltd, Novartis Pharmaceutical Company, Pfizer Inc., the Respiratory Effectiveness Group, Teva Pharmaceutical Industries, Theravance Biopharma, the UK National Health Service and Zentiva N.V.; lecture/speaking engagement fees paid to the Observational and Pragmatic Research Institute from Almirall S.A., AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Chiesi Farmaceutici S.p.A., Cipla Ltd, GlaxoSmithKline plc, KYORIN Pharmaceutical Co., Ltd, Mylan N.V., Merck & Company, Inc., Mundipharma GmbH, Novartis Pharmaceutical Company, Pfizer Inc., Skyepharma and Teva Pharmaceutical Industries; manuscript preparation fees paid to the Observational and Pragmatic Research Institute from Mundipharma GmbH and Teva Pharmaceutical Industries; travel expenses fees paid to the Observational and Pragmatic Research Institute from Aerocrine AB, AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Mundipharma GmbH, Napp Pharmaceutical Group Ltd, Novartis Pharmaceutical Company and Teva Pharmaceutical Industries; funding for patient enrolment or completion of research fees paid to the Observational and Pragmatic Research Institute from Chiesi Farmaceutici S.p.A., Novartis Pharmaceutical Company, Teva Pharmaceutical Industries and Zentiva N.V.; and payment for developing educational materials fees paid to the Observational and Pragmatic Research Institute from Mundipharma GmbH and Novartis Pharmaceutical Company. David Price is a peer reviewer for grant committees for the Efficacy and Mechanism Evaluation and Health Technology Assessment (HTA) programmes. He has stock/stock options from AKL Research and Development Ltd that produces phytopharmaceuticals, and owns 74% of the social enterprise Optimum Patient Care Ltd (in Australia, Singapore and the UK) and 74% of the Observational and Pragmatic Research Institute Pte Ltd (Singapore). Ian Pavord reports grants from GlaxoSmithKline during the conduct of the study; has received speaker’s honoraria from AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Aerocrine AB, Almirall S.A., Novartis Pharmaceutical Company and GlaxoSmithKline; has received honoraria for attending advisory board panels from Almirall S.A., AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, Dey Pharma, L.P., GlaxoSmithKline, Merck Sharp & Dohme, Schering-Plough, Novartis Pharmaceutical Company, Napp Pharmaceutical Group Ltd and RespiVert Ltd; and has received sponsorship for attending international scientific meetings from AstraZeneca plc, C.H. Boehringer Sohn AG & Ko. KG, GlaxoSmithKline and Napp Pharmaceutical Group Ltd. Mike Thomas received speaker’s honoraria for speaking at sponsored meetings or satellite symposia at conferences from Aerocrine AB, GlaxoSmithKline and Novartis Pharmaceutical Company. He has received honoraria for attending advisory panels with Aerocrine AB, Boehringer Ingelheim, GlaxoSmithKline, Merck Sharp & Dohme, Novartis Pharmaceutical Company and Pfizer Inc. during the conduct of the study. He reports grants from the National Institute for Health Research during the conduct of the study; being a member of the HTA Primary Care Community and Preventative Interventions Panel during the conduct of the study; and personal fees from GlaxoSmithKline, Novartis Pharmaceutical Company, Boehringer Ingelheim and Aerocrine AB outside the submitted work. He is a member of the British Thoracic Society/Scottish Intercollegiate Guidelines Network’s Asthma Guideline Steering Group and the National Institute for Health and Care Excellence’s Asthma Diagnosis and Monitoring Guideline Development Group. Christopher Brightling received payment via his institution of grants and personal fees from AstraZeneca plc/MedImmune, LLC, GlaxoSmithKline plc, F. Hoffmann-La Roche AG/Genentech, Inc., Novartis Pharmaceutical Company, Chiesi Farmaceutici S.p.A., Pfizer Inc., Teva Pharmaceutical Industries, Sanofi S.A./Regeneron Pharmaceuticals, Inc., Glenmark Pharmaceuticals, Mologic Ltd and Vectura Group plc.

Copyright statement

© Queen’s Printer and Controller of HMSO 2018. This work was produced by McKeever et al. under the terms of a commissioning contract issued by the Secretary of State for Health and Social Care. This issue may be freely reproduced for the purposes of private research and study and extracts (or indeed, the full report) may be included in professional journals provided that suitable acknowledgement is made and the reproduction is not associated with any form of advertising. Applications for commercial reproduction should be addressed to: NIHR Journals Library, National Institute for Health Research, Evaluation, Trials and Studies Coordinating Centre, Alpha House, University of Southampton Science Park, Southampton SO16 7NS, UK.

2018 Queen’s Printer and Controller of HMSO

Background

Asthma is one of the commonest chronic diseases in the world, affecting an estimated 300 million people. Acute exacerbations of asthma cause considerable morbidity and account for a large component of the direct and indirect costs of asthma.

Previous studies have shown that the widespread use of an asthma self-management plan can reduce exacerbations requiring oral corticosteroids and emergency health-care utilisation, as well as reduce time away from work or school because of poorly controlled asthma. However, although written self-management plans are recommended for all patients with asthma, many patients are not provided with one. Reasons for not being provided with a self-management plan include a lack of time and confusion about what to include in the plan when asthma control is deteriorating but before there is a need for oral corticosteroids.

Two large randomised, double-blind, placebo-controlled clinical trials2,3 found no benefit from doubling the dose of a patient’s usual inhaled corticosteroid2 or doubling the dose of inhaled budesonide3 when asthma control starts to deteriorate. However, other studies have suggested that higher doses (e.g. a fivefold increase4 or 1 mg of inhaled fluticasone propionate twice daily5) may be effective for the treatment of established exacerbations. A previous single-centre, randomised, double-blind, placebo-controlled clinical trial carried out by the chief investigator explored whether or not asthma exacerbations could be prevented with a self-management plan that recommended quadrupling the dose of inhaled corticosteroid at the time when asthma control starts to deteriorate (n = 403). 6 The results showed that for those participants who started on the study inhalers (n = 94) quadrupling the dose of inhaled corticosteroid led to a 36% reduction in asthma exacerbations (per-protocol analysis, p = 0.004). Unfortunately, the number of participants starting on the study inhaler varied between the two groups and the primary outcome in the intention-to-treat analysis was not significant. There is no evidence to suggest that a higher dose (i.e. a fivefold increase) is any more effective and would be associated with greater systemic activity.

In view of the limited evidence for quadrupling the dose of inhaled steroid at the point when asthma control starts to deteriorate, the UK National Institute for Health Research (NIHR) Health Technology Assessment (HTA) programme issued a funding call in February 2010 and, subsequently, commissioned the Fourfold Asthma Study (FAST) (reference number 10/143/01).

Objectives

Role of the funder

The study was funded by the NIHR HTA programme. The NIHR had input into the trial design through peer review of the proposal, but did not have a role in data collection, data analysis, data interpretation or the writing of the final report. The corresponding author had access to all the data and was responsible for the decision to submit the final report.

Trial design

The FAST was a multicentre, pragmatic, normal care-controlled, randomised controlled trial (RCT) of 12 months’ duration (Figure 1). Adult asthma patients were randomised (1 : 1) to one of two asthma self-management plans: usual care or modified. Both self-management plans were identical at zones 1, 3 and 4, but at zone 2 (worsening of asthma symptoms) the usual-care group was advised to follow the current guidelines of increasing bronchodilator medication when asthma control begins to deteriorate. The modified self-management plan advised participants to increase bronchodilator medication and quadruple their inhaled corticosteroid dose. Both self-management plans advised participants to increase their medication for a maximum of 14 days, or for a shorter duration if asthma symptoms started to improve, before returning to their normal treatment, which was actively promoted in both self-management plans.

FIGURE 1.

Trial visit flow chart.

Participants were expected to attend three scheduled visits: baseline, 6 months and 12 months. These visits were conducted at the participants’ local general practitioner (GP)’s clinic or hospital (or via the telephone if easier for the participant). During these visits the research nurse reviewed the participant’s diary card to assess their adherence to the self-management plan and use of inhaled steroids, reviewed the participant’s asthma control to determine if there had been any unreported activation of zones 2, 3 or 4 since the previous visit and asked participants to complete the EuroQol-5 Dimensions, three-level version (EQ-5D-3L), questionnaire and Service Use Questionnaire (see Appendix 1).

In addition to the scheduled visits, participants were expected to attend post-activation visits no less than 14 days after activating zone 2 of the asthma self-management plan; for some participants multiple visits took place in the 12-month participation period. Prior to the visit participants were expected to complete a diary card. For FAST two diary cards were used: one diary card for those participants who used corticosteroid inhalers and a second for those on combination inhalers. The participants used the diary cards to record peak flow, to record the asthma medication used to manage symptoms (including the number of puffs of usual preventer inhaler, of extra corticosteroid inhaler and of reliever inhaler and whether or not any systemic corticosteroids were taken) and to document whether they had an asthma-related GP or hospital appointment. The diary cards acted as an aide-memoire during the post-activation visit; the research nurse reviewed the daily peak flow measurements, any health-care consultations attended, the use of inhaled corticosteroids and systemic corticosteroids, ascertained if any adverse events (AEs) had occurred and assessed the participant’s adherence to the asthma self-management plan. Participants were also asked to complete the Juniper et al. Mini Asthma Quality of Life Questionnaire (Mini AQLQ). 8 The study assessments are outlined in Table 1.

In addition, it was planned that the first 200 participants recruited at the Nottingham and Liverpool sites would have the option (if they agreed) to have their inhaled corticosteroid inhaler fitted with a smart inhaler electronic dose counter for adherence purposes. The main purpose of this was to compare an electronic record of inhaler use with the participants’ self-reported inhaler use and, therefore, the overall adherence to their allocated self-management plan. The smart inhaler was planned to be used to independently validate the accuracy of the participants’ self-reported adherence.

The implementation of the smart inhaler was challenging and, unfortunately, some technical issues (the smart inhaler not working correctly and a short battery life) prevented the implementation of the initial batch of devices into the trial.

Following a period of uncertainty around delivery and the reliability of new devices to replace those that did not function accurately, it was the unanimous view of the Trial Steering Committee (TSC), at a meeting held in July 2014, that it was not practical to pursue further use of monitoring devices as the chance of any informative data being obtained from them prior to the end of the study was low. As a result, the planned interim analysis to determine whether or not self-reported adherence to adjustments to inhaled steroid dose was similar to that captured by the electronic devices was not conducted.

Recruiting centres

Recruitment took place in both primary and secondary care across England and Scotland. There were 10 secondary care sites: Nottingham City Hospital, Leicester Glenfield Hospital, Freeman Hospital Newcastle, Aintree University Hospital, Aberdeen Royal Infirmary, Royal Liverpool University Hospital, King’s Mill Hospital, Arrowe Park Hospital, Blackpool Victoria Hospital and Bradford Royal Infirmary. In addition, there were 171 primary care Research Initiative Sites (RISs) across 11 CRN regions: North East and North Cumbria, North West Coast, Greater Manchester, East Midlands, West Midlands, West of England, Thames Valley and South Midlands, Eastern, Kent, Surrey, Sussex, Wessex and South West Peninsula.

Participants were identified through secondary care and primary care. In secondary care, participants were identified from patients attending respiratory outpatient appointments at the individual recruiting centres and also through running database searches of participants who had previously participated in asthma studies and had given consent to be contacted again for future studies. In Scottish centres, the Scottish Primary Care Research Network identified potential participants in primary care who were subsequently recruited at local secondary care sites. In primary care, local CRNs liaised directly with GP practices that acted as RISs that performed a database search to identify potential participants. Potentially eligible participants were sent a participation invitation pack that included an invitation letter flyer about the trial and, in some practices, a copy of the Participant Information Sheet.

A local press release was issued at the start of the trial. Posters and flyers were displayed in recruiting centres and, where possible, a digital flyer was displayed in recruiting centre waiting areas. Posters were also displayed in pharmacies in the Nottinghamshire area and information flyers were placed in the bags of patients who were collecting prescriptions for asthma medication. The trial was also promoted online by Asthma UK.

General practitioner surgeries local to the recruiting hospitals were used as Participant Information Centres (PICs) by displaying posters and flyers.

Participants

Patients were considered eligible for entry into the trial if the following inclusion criteria were met:

• men or women aged ≥ 16 years

• clinician-diagnosed asthma treated with a licensed dose of inhaled corticosteroid [i.e. steps 2–4 of the British Thoracic Society/Scottish Intercollegiate Guidelines Network (BTS/SIGN) guidelines]2

• one or more exacerbations in the last 12 months requiring treatment with systemic corticosteroids

• current smokers could be included provided that the recruiting centres had good evidence of underlying asthma (i.e. a life-long history of asthma, a > 12% forced expiratory volume in 1 second (FEV₁) reversibility, or sputum or blood eosinophilia).

In addition, patients were not entered into the trial if any of the following exclusions applied:

• a history more in keeping with smoking-related chronic obstructive pulmonary disease (COPD) (i.e. smoked > 20 pack-years, without evidence of significant reversibility or eosinophilia)

• on maintenance systemic corticosteroids (i.e. step 5 of the BTS/SIGN guidelines2)

• using a combination inhaler for both maintenance and relief treatment

• experienced an exacerbation within 4 weeks of randomisation

• women who were pregnant, breastfeeding or who were planning to become pregnant.

Interventions

Outcome measures

Data collection

Trial data generated by all centres were entered by site staff directly into a web-based bespoke database, designed and maintained by the Nottingham Clinical Trials Unit (NCTU). Access to the trial database was controlled by user logins, and users could enter or edit data only for their regional centre. Participant questionnaires, completed at clinic visits, were entered into the trial database by site staff. If participants had not activated zone 2 of their self-management plan and were unable to attend the 6- and/or 12-month visit in person, site staff contacted participants via telephone and completed the trial database directly with the information provided to them over the telephone. Any missing and/or ambiguous data were queried with site staff, by the NCTU team, and resolved wherever possible.

Participants were asked to complete their diary cards prior to attending their post-activation visit. This diary card is where the participants recorded peak flow and the asthma medication used to manage symptoms (including the number of puffs of usual preventer inhaler, extra corticosteroid inhaler, reliever inhaler and whether or not any systemic corticosteroids were taken) and documented whether they had an asthma-related GP or hospital appointment. The diary was reviewed by the site staff at each clinic appointment and this information was used to complete the relevant sections of the trial database.

Site staff were also required to assess the participant’s adherence to the asthma self-management plan by reviewing the diary card and discussing, during the visit, whether or not, and how, participants changed their inhaled treatment since activating zone 2 of their self-management plan. Site staff completed the adherence assessment directly into the web-based bespoke system. Initially, there was the option for the first 200 participants recruited to measure their actuation adherence with the smart inhaler. The participant’s corticosteroid inhaler would be fitted with the electronic dose counter to record the date and time of each actuation. The data would then be directly downloaded from the device during the participant’s visit and uploaded to a separate database. The smart inhaler devices were not implemented because of the challenges described in Trial design and all adherence data were assessed by the site staff.

The web-based bespoke system generated automated notification e-mails for sites to remind them of upcoming and overdue 6- and 12-month visits. These notification e-mails were sent to sites on a monthly basis on the first day of every month during the recruitment and follow-up phase of the trial.

Those participants who provided additional consent, were sent monthly text reminders to remind them to adhere to their asthma self-management plan. The text message service was automated, with text messages automatically initiated on the first day of every month at approximately noon. To maintain confidentiality the participant’s mobile phone number was entered on to the database by site staff at the point of consent. These data were then encrypted and stored confidentially. Only research staff at their own specific site had access to their participants’ mobile phone numbers. If the participant withdrew, was lost to follow-up or had completed the trial, then the text messages automatically stopped being sent to the participant.

Sample size

A reduction of one-third in the number of people requiring treatment with oral corticosteroids was considered an important treatment effect by a group of local GPs, asthma nurses and asthma experts.

With 1000 participants per group, a log-rank test (at the two-sided 5% significance level) provided at least 90% power to detect a difference of 30% (relative effect), assuming an exacerbation rate of 13% in the control group. A 13% exacerbation rate requiring systemic corticosteroids was the lowest level seen in the control group of previous studies of this type6,10 and so provided a conservative estimate. The study initially proposed to recruit 2300 participants to allow for participants lost to follow-up (i.e. approximately 15% lost to follow-up). The study was not powered for the subgroup analysis performed on smoking status or dose of maintenance inhaled steroid dose at baseline.

The power calculation was revised in March 2015 in consultation with the NIHR HTA programme. The overall event rate in the first 226 participants recruited was higher (around 50%) in those reaching the 12-month follow-up time point. The exacerbation rate at this time was thought to be high because of more participants being recruited from secondary care and, therefore, having more severe asthma; hence, a lower exacerbation rate was used to revise the sample size calculation. Assuming an exacerbation rate in the control group of 17% and 90% power, and still estimating a one-third reduction in the fourfold increase group, then a sample size of 1542 participants was needed for analyses. Allowing for 20% of participants being lost to follow-up, the study aimed to recruit between 1774 and 1850 participants before the close of recruitment on 31 January 2016.

For those participants who were lost to follow-up, where possible, site staff reviewed participant computerised medical records to document if they had had an asthma-related GP appointment, if their asthma had exacerbated or if they had been prescribed systemic steroids.

Stopping rules and discontinuation

Ongoing adherence to the self-management plans was assessed by the Data Monitoring Committee in accordance with the criteria outlined in Table 4.

Randomisation

Randomisation was stratified by regional centre (Box 1), smoking status (yes/no) and maintenance inhaled corticosteroid dose (high/low dose; Table 5). Participants were allocated with equal probability to the two trial treatment groups. Recruiting sites were grouped into regional centres, which grouped practices to the appropriate CRN regions or secondary hospital care sites (Box 1). The treatment group to which a participant was assigned was determined by a computer-generated pseudo-random code, with random permuted blocks of randomly varying size, that was created by the NCTU in accordance with its standard operating procedure. The data were held on a secure University of Nottingham server.

Research nurses accessed the randomisation website by means of a remote, internet-based randomisation system developed, and maintained, by NCTU. Access was controlled by unique user logins. The sequence of treatment allocations was concealed until interventions had all been assigned and recruitment and data collection were complete. The chief investigator, trial team and trial statisticians were blinded to treatment allocations until the database was locked.

Blinding

Owing to the nature of the self-management plans allocated in the trial, it was not possible to blind site staff or participants to their treatment allocation. Efforts were made to minimise the expectation bias by detailing in the trial documents that the evidence supporting the quadrupling of the inhaled dose of corticosteroid at the time of worsening asthma symptoms was limited, and it was not yet known whether or not the intervention offered any benefit over usual care. Both groups were also provided with tailored self-management plans that were explained to them in detail, ensuring that both groups received similar instruction on how to best use their asthma self-management plan.

Throughout the trial, prior to database lock, the blinding allocation was preserved for the chief Investigator, trial statisticians, the trial team and TSC members.

Full details of blinding arrangements are summarised in Table 6.

Statistical analysis

Analyses are detailed in the statistical analysis plan (www.nottingham.ac.uk/nctu/trials/respiratory.aspx#FAST), which was finalised prior to database lock and release of the treatment allocation codes for analysis.

All analyses were based on the intention-to-treat principle, for example analysed as randomised regardless of adherence to self-management plan. All analyses were carried out using Stata^®/SE 13.1 (StataCorp LP, College Station, TX, USA).

Summary of changes to the protocol

The full protocol and statistical analysis plan are available on the NCTU’s website (www.nottingham.ac.uk/nctu/trials/respiratory.aspx#FAST). A summary of changes made to the protocol after the start of recruitment is listed in Appendix 2.

Recruitment

Recruitment to the trial took place between 17 May 2013 and 29 January 2016 (Figure 2).

FIGURE 2.

Cumulative recruitment.

During this time 20,695 asthma patients were contacted and invited to take part in the study (patient contact data were not provided by 77 of the initiated sites). Of these, 4811 were assessed for eligibility and 1922 were subsequently randomised (Figure 3). Of the 2889 participants who were screened but not randomised, 860 (30%) declined to participate and 2029 (70%) failed to meet the eligibility criteria (Table 7).

FIGURE 3.

The Consolidated Standards of Reporting Trials (CONSORT) participant flow diagram. a, Invited figure reflective of the data provided by 132 initiated sites at the time of reporting.

Initially, recruitment was slow, and after 6 months was only 25% of the target as a result of a combination of delays with contracting and a higher than expected rate of non-eligibility. Following review of site recruitment and recruitment trends in primary and secondary care, focus was shifted to concentrate on primary care RISs. Although the RISs were initially achieving target after around 6 months, the pool of potential participants was exhausted and recruitment dropped. In February 2014, FAST’s Trial Management Group met the trial funders to agree on a strategy that new RISs should be opened to replace sites as they became inactive. Following the meeting, an ambitious initiation plan was undertaken, with 54 sites opening between March and December 2014 and a further 107 sites opening in 2015.

In order to meet recruitment targets the NIHR HTA programme agreed to an 11-month recruitment extension (January 2015). The trial completed recruitment on 31 January 2016 with 1922 randomised participants and 196 RISs and 11 secondary care sites initiated throughout the recruitment period. At the outset of the study it was envisaged that approximately 80% of participants would be recruited from primary care (i.e. PICs). In total, 81% of participants were recruited from primary care (63% RISs and 18% PICs) and 19% from secondary care sites. The number of participants randomised to each group was well balanced across regional randomisation centres (Table 8).

Baseline data

Follow-up

Primary and secondary outcomes

Primary outcome: time to first asthma exacerbation

Primary analysis

The number of participants having an exacerbation of asthma in the year after randomisation was 484 (51.6%) in the usual-care group and 420 (45.0%) in the modified self-management group (Figure 4). Kaplan–Meier curves for the time to first asthma exacerbation are shown in Figure 4. The adjusted hazard ratio for the time to first asthma exacerbation in the modified self-management group compared with the usual-care group was 0.81 (95% CI 0.71 to 0.92; p = 0.002; Table 16).

FIGURE 4.

Kaplan–Meier curves for the time to first asthma exacerbation by allocated group.

TABLE 16 - Primary outcome: time to first asthma exacerbation

The adjusted hazard ratio was estimated using the Cox proportional hazards model adjusted for randomisation stratification variables [smoking status and dose of inhaled steroids (high/low), with shared frailty for regional randomisation centre]. The unadjusted hazard ratio was 0.82 (95% CI 0.72 to 0.94).

Note

Follow-up time is the time to first exacerbation or the time in trial if there was no exacerbation.

Primary outcome	Intervention arm		Adjusted hazard ratioa (95% CI); p-value
	Usual care (N = 938)	Modified (N = 933)
Total number (% of total) with exacerbation	484 (51.6%)	420 (45.0%)
Total follow-up time (person-years)	610.3	649.8
Rate (per person-years)	0.79	0.65	0.81 (0.71 to 0.92); p = 0.002

Secondary outcomes

Unscheduled health-care consultations and the use of systemic corticosteroids for asthma

Time to first use of systemic corticosteroids for asthma and time to first unscheduled health consultation for asthma

Figure 5 shows the Kaplan–Meier curves for the time to first use of systemic corticosteroids and the time to first unscheduled health-care consultation is shown in Figure 6. The adjusted hazards ratios are 0.76 (95% CI 0.65 to 0.88; p < 0.001) for the use of systemic corticosteroids and 0.82 (95% CI 0.70 to 0.92; p = 0.002) for unscheduled health-care consultations.

FIGURE 5.

Kaplan–Meier curves for the time to first requiring systemic corticosteroids for asthma by allocated group.

FIGURE 6.

Kaplan–Meier curves for the time to first unscheduled health-care consultation for asthma by allocated group.

Total number of courses of systemic corticosteroids for asthma, unscheduled health consultations and exacerbations for asthma

Table 19 shows that the number of participants using systemic corticosteroids, having an unscheduled health-care consultation and an exacerbation (systemic corticosteroids or unscheduled health-care consultation), was lower in the modified self-management group than in the usual-care group.

TABLE 19 - Summary of unscheduled health-care consultations and use of systemic corticosteroids for asthma

Adjusted for randomisation stratification variables (smoking status, dose of inhaled steroids and regional centre). All analysis models include 1874 participants.

Not known for use of systemic corticosteroids if eCRF page on oral steroid use for asthma reported as unknown or not completed; not known for unscheduled health-care consultation if eCRF page for health-care consultations for advice on worsening of asthma symptoms completed as unknown or not completed; and not known for exacerbations if either eCRF page completed as unknown or not completed and no exacerbation reported on other page or both pages were completed as unknown or not completed. These participants are included in the analysis models as having no courses of steroids/health-care consultations/exacerbations (as appropriate) for consistency with the primary analysis where these participants are censored.

Notes

For participants who did not complete follow-up, information derived up to the time they exited from the trial.

This table also includes data for participants who exacerbated on the same day as randomisation and so were not able to be included in the analysis of time to first exacerbation.

Secondary outcome	Intervention arm		Adjusted intervention effecta (95% CI)
	Usual care (N = 939)	Modified (N = 935)
Use of systemic corticosteroids
Any courses, n (% of total)
No	552 (59)	614 (66)	Risk difference –7.0% (–11.3% to –2.7%)	Risk ratio 0.83 (0.74 to 0.93)
Yes	377 (40)	311 (33)
Not knownb	10 (1)	10 (1)
Total number of courses	n = 929	n = 925	Incidence rate ratio 0.82 (0.70 to 0.96)
Mean (SD)	0.61 (0.93)	0.5 (0.86)
1, n (% of total)	254 (27)	212 (23)
2, n (% of total)	82 (9)	68 (7)
3 or more, n (% of total)	41 (4)	31 (3)
Any unscheduled health-care consultations
Any, n (% of total)
No	490 (52)	543 (58)	Risk difference –6.8% (–11.1% to –2.4%)	Risk ratio 0.86 (0.78 to 0.95)
Yes	442 (47)	379 (41)
Not knownb	7 (1)	13 (1)
Total number of unscheduled health-care consultations	n = 932	n = 922	Incidence rate ratio 0.86 (0.75 to 0.99)
Mean (SD)	0.84 (1.23)	0.73 (1.19)
1, n (% of total)	261 (28)	224 (24)
2, n (% of total)	96 (10)	83 (9)
3 or more, n (% of total)	85 (9)	72 (8)
Exacerbation: use of systemic corticosteroids and/or unscheduled health-care consultation for asthma
Any exacerbations, n (% of total)
No	445 (47)	499 (53)	Risk difference –6.7% (–11.2% to –2.3%)	Risk ratio 0.87 (0.80 to 0.95)
Yes	485 (52)	422 (45)
Not knownb	9 (1)	14 (1)
Total number of exacerbations	n = 930	n = 921	Incidence rate ratio 0.88 (0.77 to 1.01)
Mean (SD)	0.95 (1.29)	0.84 (1.26)
1, n (% of total)	270 (29)	235 (25)
2, n (% of total)	119 (13)	97 (10)
3 or more, n (% of total)	96 (10)	90 (10)

Similarly, the total number of courses of systemic corticosteroids, unscheduled health-care consultations and exacerbations per participant was lower in the modified self-management group than in the usual-care group (Table 19).

Serious adverse events

Serious adverse events were reported during the 14-day period following activation of zone 2 (or above) of the self-management plan. In addition, diagnoses of pneumonia up to 1 month after the 14-day activation period of zone 2 or above were also considered to be SAEs.

A total of 22 (4%) participants in the usual-care group and 11 (2%) participants in the modified self-management group who activated zone 2 or above had at least one SAE (Table 25).

Eighteen of the 32 SAEs in the usual-care group were attributable to hospitalisations for asthma, compared with 3 of the 11 SAEs in the modified self-management group.

There were eight events in the usual-care group and six events in the modified self-management group relating to pneumonia, lower respiratory tract infections and influenza (Table 25).

One participant in the modified self-management group died after severe pneumonia. This event was not classified as related to trial treatment.

One of the 11 SAEs in the modified self-management group was classified as possibly related to trial treatment. This participant had pneumonia (outside the 14-day period following activation to zone 2) and had fourfolded their usual medication (4000 µg) on two occasions (once for 7 days and one for 6 days) prior to the event.

Non-serious adverse events

Non-serious AEs were reported during the 14 days following activation of zone 2 of the self-management plan. Only adverse events that are known side-effects of inhaled corticosteroids, such as oral candidiasis (i.e. thrush) and dysphonia (i.e. hoarseness), were intended to be collected from discussion with the participant and information from the diary card.

Ten (2%) participants in the usual-care group and 41 (7%) participants in the modified self-management group who activated zone 2 or above had at least one non-serious AE (Table 26). Of the 56 non-serious AEs in the modified self-management group, 44 were classified as definitely or probably related to inhaled corticosteroids, compared with 6 of the 13 non-serious adverse events in the usual-care group. The breakdown of the type of events is shown in Table 26.

Introduction

The cost-effectiveness analysis (CEA) was conducted alongside the FAST to establish the value for money of temporarily quadrupling the dose of inhaled corticosteroid compared with usual care.

The objectives of the CEA were to:

1. identify the related costs associated with delivering the treatments

2. measure the participants’ use of respiratory-related health and social care services

3. compare the estimated mean cost per participant between the two intervention groups

4. estimate the health benefits of the trial interventions using QALYs calculated from the EQ-5D-3L questionnaire and the number of exacerbations prevented

5. compare the cost difference between the two intervention groups with difference in effectiveness and generate incremental cost-effectiveness ratios (ICERs)

6. test the uncertainty of the calculated ICERs, using the bootstrapping method, and generate cost-effectiveness acceptability curves (CEACs) to demonstrate the probability of the modified treatment being cost-effective over and above usual care.

Methods

Following the National Institute for Health and Care Excellence (NICE)’s Guide to the Methods of Technology Appraisal 2013,15 the analysis was conducted from the NHS/Personal Social Services perspective, with costs expressed in Great British pounds (£) for the financial year 2014–15. 15 A total of 1922 participants were randomised; however, 51 of these participants were excluded as they had no information from randomisation. The follow-up for the analysis was 6 months and 12 months from randomisation. All costs were inflated to 2014–15 price levels, where necessary, using the Hospital and Community Health Services pay and price inflation index. 16,17 No discount rate was applied as the follow-up was 12 months.

Results

A total of 1922 participants were recruited to the trial, with 51 excluded, leaving 1871 participants analysed in the economic evaluation (935 in the modified self-management group and 939 in the usual-care group). The base-case CEA was based on a MI data set, in which all the missing values were imputed using the MI method.

Cost-effectiveness analysis and uncertainty

Two sets of cost-effectiveness analyses were conducted using QALYs and the number of exacerbations per participant. As the modified treatment was both less costly and more effective for both health outcomes, the modified treatment is said to be ‘dominant’. However, Table 31 reflects the uncertainty with this result, as shown by the 95% bootstrapped CIs.

TABLE 31 - Results of CEA

CEA results	Intervention arm		Difference in mean (bootstrapped 95% CI)
	Usual care (N = 939)	Modified (N = 935)
Total cost, mean (SE)	£431 (£43)	£415 (£42)	–£24 (–£122 to £71)
QALY, mean (SE)	0.74 (0.01)	0.76 (0.09)	0.02 (–0.00 to 0.04)
Exacerbations, mean (SE)	0.95 (0.04)	0.84 (0.04)	–0.10 (–0.21 to 0.01)
ICER
QALY (bootstrapped 95% CI)	Dominant (–£21,699 to £16,268)
Exacerbations (bootstrapped 95% CI)	Dominant (–£1999 to £2492)

However, the difference between costs, QALYs and exacerbations was not statistically significant. This suggested that there is significant uncertainty surrounding these estimates. To investigate this, a non-parametric bootstrapping technique was investigated. A cost-effectiveness scatterplot was produced from the bootstrapping results for difference in QALYs and difference in cost and then, again, for difference in exacerbations and difference in cost. The results of the 5000 re-samples for each outcome were plotted on a cost-effectiveness plane (Figure 7), visually displaying any uncertainty surrounding the mean differences in costs and benefits between the intervention and usual-care groups. The CEP in Figure 7a shows the majority of the plots falling in the south-east quadrant (65%), suggesting greater QALYs and lower costs for the modified self-management group. With 29% of the plots falling in the north-east quadrant, there is some uncertainty surrounding the costs; however, the majority of plots fall below both the £30,000- and £20,000-threshold line, implying a high probability of cost-effectiveness for the modified self-management group.

FIGURE 7.

Cost-effectiveness planes (base case).

Figure 7b shows the majority of the plots falling in the west quadrants, suggesting a positive health gain in terms of exacerbations prevented, but also uncertainty in terms of costs with plots falling in both the north and south quadrants.

Using the bootstrapped replicates we also generated a CEAC (Figure 8), which provides a plot of probabilities that the intervention was cost-effective (y-axis) against all potential values of willingness-to-pay thresholds (x-axis). This can be generated only for the QALY outcome, as there is not a threshold for exacerbations prevented. The CEAC (Figure 8) shows that, with a willingness-to-pay threshold of £20,000, there is a 94% chance of the intervention being cost-effective.

FIGURE 8.

Cost-effectiveness acceptability curve.

Complete-case analysis (sensitivity analysis)

In order to explore the potential impact of missing data on the results, a sensitivity analysis was conducted using complete cases. Complete costs and QALYs were available for 1041 participants and complete costs and the number of exacerbations were available for 1047 participants.

Table 32 shows the complete-case analysis results for complete QALYs and costs. As with the primary analysis, the results were cost-saving (–£109, bootstrapped 95% CI –£259 to £41; p = 0.148) and not statistically significant. The results of the QALYs showed a lower difference, of 0.01 (bootstrapped 95% CI –0.02 to 0.04; p = 0.517), compared with 0.02 in the primary analysis. However, the results did not reach statistical significance.

TABLE 32 - Complete-case analysis: QALYs

Includes cost of death.

Costs and QALYs	Intervention arm
	Usual care (N = 512)	Modified (N = 529)
Intervention, mean (SD)	£22 (£34)	£51 (£74)
Resource use at 6 months, mean (SD)	£240 (£766)	£139 (£275)
Resource use at 12 months, mean (SD)	£220 (£920)	£191 (£853)
Total service use,a mean (SD)	£461 (£1459)	£333 (£930)
Total cost, mean (SD)	£483 (£1464)	£384 (£943)
Adjusted difference (bootstrapped 95% CI)	–£109 (–£259 to £41); p = 0.148
EQ-5D-3L scores, mean (SD)
Baseline	0.80 (0.26)	0.82 (0.23)
6-month follow-up	0.80 (0.26)	0.82 (0.23)
12-month follow-up	0.81 (0.26)	0.81 (0.24)
QALYs, mean (SD)	0.80 (0.23)	0.82 (0.21)
Adjusted difference (bootstrapped 95% CI)	0.01 (–0.02 to 0.04); p = 0.517

Table 33 shows the complete-case results for the number of exacerbations. As with the complete-case results for the QALYs, the costs for the complete exacerbations results were cost-saving, with an adjusted difference of –£110 (bootstrapped 95% CI –£265 to £39; p = 0.175). The complete-case results for the number of exacerbations showed a greater difference in the number of exacerbations, with an adjusted difference of –0.18 (bootstrapped 95% CI –0.34 to –0.02) and the results did reach statistical significance (p = 0.046).

TABLE 33 - Complete-case analysis: exacerbations

Includes cost of death.

Costs and exacerbations	Intervention arm
	Usual care (N = 514)	Modified (N = 533)
Intervention, mean (SD)	£23 (£34)	£50 (£74)
Resource use at 6 months, mean (SD)	£239 (£765)	£138 (£274)
Resource use at 12 months, mean (SD)	£219 (£918)	£191 (£850)
Total service use,a mean (SD)	£459 (£1457)	£333 (£927)
Total cost, mean (SD)	£482 (£1461)	£383 (£940)
Adjusted difference (bootstrapped 95% CI)	–£110 (–£265 to £39); p = 0.144
Exacerbations, mean (SD)	1.07 (1.34)	0.90 (1.30)
Adjusted difference (bootstrapped 95% CI)	–0.18 (–0.34 to –0.02); p = 0.046

The modified self-management group had lower costs and better health outcomes; therefore, the modified treatment was said to be ‘dominant’. However, the cost and QALY difference were not statistically significant, so CEPs were used to explore this uncertainty. Figure 9 shows the CEP for QALYs and for exacerbations on the left-hand side and for exacerbations on the right-hand side. The majority of the plots in the CEP in Figure 9a fall in the south-east quadrant (71%), inferring lower costs and greater QALYs for the modified self-management group. However, compared with the base case, more plots fall in the south-west quadrant (21% vs. 4%), where both incremental costs and effects are negative. The majority of plots fall below both the £30,000- and £20,000-threshold line, suggesting a high probability of cost-effectiveness. The CEP in Figure 9b shows certainty in reducing exacerbations, with the majority of the plots falling in the west quadrants, which was also reflected in the significant p-values. The CEAC in Figure 10 shows an 86% probability of the intervention being cost-effective at a threshold of £20,000.

FIGURE 9.

Cost-effectiveness planes (complete case).

FIGURE 10.

Cost-effectiveness acceptability curve (complete case) for 12 months.

Summary/conclusion

This economic evaluation assessed the cost-effectiveness of temporarily quadrupling the dose of inhaled corticosteroid compared with usual care.

The mean intervention cost was £42 (SE £2) per participant in the modified self-management group and £17 (SE £1) per participant in the usual-care group. Taking into consideration the wider respiratory-related health-care resource use, participants who received the modified treatment had non-significantly lower total mean costs over the 12-month period after adjusting for covariates. The adjusted total cost difference was –£24 (bootstrapped 95% CI –£122 to £71).

It is recommended by NICE that cost-effectiveness be expressed in terms of cost per QALY. In this study, there was a non-significant trend towards higher QALYs associated with the modified treatment (adjusted difference of 0.02, bootstrapped 95% CI –0.005 to 0.04). The economic evaluation also used the number of exacerbations. There was, once again, a non-significant trend towards fewer exacerbations associated with the modified treatment (adjusted difference of –0.10, bootstrapped 95% CI –0.22 to 0.01).

The base-case results were based on a MI data set. As the modified treatment was both more effective and less costly, it was said to be ‘dominant’ in terms of both QALYs and the number of exacerbations prevented. Although the differences in QALYs and costs were not statistically significant, the CEAC demonstrated a 94% probability of it being cost-effective at a £20,000 threshold.

A sensitivity analysis was carried out using complete cases to explore the impact of missing data. The results showed a lower cost for the modified self-management group and higher cost for the usual-care group, resulting in the modified treatment still being cost-saving for both the exacerbation complete-case analysis (adjusted difference of –£110, bootstrapped 95% CI –£265 to £39) and the QALY complete-case analysis (–£109, bootstrapped 95% CI –£259 to £41). However, the difference in QALYs was reduced, resulting in the modified self-management group being less effective than in the base case (adjusted difference of 0.01, bootstrapped 95% CI –0.02 to 0.04). On the other hand, the difference in exacerbations was greater and statistically significant (adjusted difference of –0.18, bootstrapped 95% CI –0.34 to –0.02), unlike the base-case analysis. As both health outcomes saw greater improvement and lower costs than usual care, the modified treatment was said to be ‘dominant’. However, this should be interpreted with caution as the cost difference did not reach statistical significance and the CEPs demonstrate the uncertainty surrounding this result.

Including reliever inhaler costs within the intervention cost made little difference to the results, as these costs were so small. Changing the leeway of the questionnaires to 1 month, from 2 months, decreased the adjusted total cost difference to –£10 (bootstrapped 95% CI –£94 to £74) and reduced the QALY difference to 0.01 (bootstrapped 95% CI –0.01 to 0.03). However, changing the leeway of the questionnaires did not change the direction of the results and it still had an 87% chance of being cost-effective at a threshold of £20,000.

An exploration of wider societal costs was carried out with the available data. On the whole, there was little difference between the intervention groups, with slightly greater costs seen for the modified self-management group. However, some of this greater cost was due to one participant reporting very high travel costs. Although conclusions could not be drawn because of missing data, the exploration suggested that there may be costs incurred by participants as a result of hospital and GP visits.

In conclusion, the economic evaluation of the FAST has provided us with evidence showing that quadrupling the inhaled corticosteroid use during the activation zone results in better clinical outcomes. Moreover, the CEA shows that this intervention is likely to be a cost-effective intervention in comparison with usual care.

Summary/conclusion

Our results demonstrate that a self-management plan that recommends a temporary quadrupling of inhaled corticosteroids at the time of deteriorating asthma control can prevent asthma exacerbations, compared with the usual self-management plan. This supports the findings from a previous trial. 6

Overall, 51.6% of the usual-care self-management group had an asthma exacerbation (defined as the need for systemic corticosteroids and/or unscheduled health-care consultation) at least once in the 12-month study period, compared with 45% of the modified self-management group, with an adjusted hazard ratio for time to first exacerbation of 0.81 (95% CI 0.71 to 0.92; p = 0.02). This equates to 15 patients needing to be given the modified self-management plan for one additional patient to benefit (i.e. avoid an asthma exacerbation, 95% CI 9 to 43 patients). There was no evidence in the subgroup analysis that the intervention effect differed for the time to first exacerbation according to inhaled corticosteroid dose or smoking status.

Each of the 12 secondary outcomes favoured the modified treatment group. There were fewer participants who were prescribed systemic corticosteroids and attended unscheduled health-care consultations in the modified self-management group than in the usual-care self-management group. However, there was a large number of missing data for some secondary outcomes (i.e. peak flow and the Mini AQLQ) which limited confidence in the findings. However, the results were in keeping with the primary outcome and secondary outcomes for which data completion was greater, providing some reassurance around their reliability.

The safety data continued to support the clinical benefit of a temporary fourfold increase in of inhaled corticosteroids as participants in the modified treatment group reported fewer asthma-related hospitalisations (three in the modified self-management group compared with 18 in the usual-care group). The modified self-management group did experience a higher frequency of treatment-related side effects of inhaled corticosteroids, such as oral thrush (56 events reported by 41 participants in the modified self-management group and 13 events reported by 10 participants in the usual-care group), but this was expected because of the nature of the intervention. Local adverse effects such as these are not usually a major problem and are usually easily treated with local therapy. Of more concern are reports of pneumonia, especially in patients with COPD, and adverse effects related to systemic absorption, such as adrenal suppression, osteoporosis and cataract. The median dose of inhaled corticosteroid in our study was 800 mcg/day, so quadrupling this would equate to the equivalent of 3200 mcg/day of inhaled beclometasone. Unfortunately, the systemic effects of high-dose inhaled corticosteroids are not well described and the dose potency in terms of prednisolone equivalents appears to vary from tissue to tissue. 31 In terms of milligrams of prednisolone, adrenal suppression from 1.5 mg/day of fluticasone (approximately 3 mg of beclometasone) has been estimated to have approximately the same effects on morning cortisol suppression as between 10 and 20 mg of prednisolone. 31 As the study included patients on 2000 mcg of inhaled fluticasone, and if the dose potency ratio between fluticasone propionate and prednisone is linear, then the quadrupled dose could have the same systemic effects as a course of prednisolone used to treat asthma exacerbations.

Finally, although there have been reports of an increased risk of respiratory infections in patients on long-term high-dose inhaled steroids,32 the study found no evidence of an increased incidence of pneumonia.

Health economics

To date, FAST is the first large randomised controlled trial that has assessed the cost-effectiveness of temporarily quadrupling the dose of inhaled steroid to reduce asthma exacerbations.

When using QALYs as a standard health outcome measurement, the modified treatment was said to be ‘dominant’. When using the NICE decision threshold of £20,000–30,000, the modified treatment had a 94–95% probability of being cost-effective. The complete-case analysis also showed a 86% probability of the modified treatment being cost-effective at a threshold of £20,000; however, this probability decreased as the threshold increased. This is because the complete-case results had more plots falling into the south-west quadrant (i.e. less costly and less effective). The CEAC assumes that the amount saved in order to give up one QALY increases with the threshold. Therefore, more plots were excluded in the south-west quadrant as this threshold increased. This led to a lower probability of cost-effectiveness at higher thresholds, as many plots fell into the south-west quadrant.

When using the number of exacerbations as the health outcome, the modified treatment was once again ‘dominant’. However, as no decision-making threshold exists, the probability of cost-effectiveness could not be assessed.

Reducing the number of exacerbations per participant should lead to fewer hospitalisations, which account for the greatest cost of wider health care. A lower number of hospitalisations was reported at 6 months for the modified self-management group and was reflected in the much lower cost at this follow-up; however, this did not persist through to the 12-month follow-up and the overall difference in resource-use costs was not statistically significant. This may suggest that the overall reduction in the number of exacerbations was not great enough to make a significant impact on hospital admissions, but over a longer period of time this may accumulate a greater cost-saving. This may have also been impacted by an imbalance at baseline, but as the study did not collect baseline costs the analyses still need to be interpreted with caution.

The strength of the economic analysis has been impacted by a few limitations of the study. First, there was no baseline service-use questionnaire. Although the randomisation ensures that characteristics and confounders are balanced between the groups at baseline, it may not have been balanced for health-care service use. The questionnaires were also not always delivered at the right time, and so assumptions had to be made as to what was considered an appropriate time frame and data were lost because of this. For the intervention costing, only what was reported on the diary cards could be used. It is, therefore, possible that some participants may have activated but not recorded their inhaler use and the study would not have been able to cost this. Once more, nine participants in the usual-care group were reported being given extra steroid inhalers and self-reported using them on their diary cards, so there might have been some crossover between groups.

Relevance to existing literature

In June 2016, the Cochrane review relating to whether or not an increase in corticosteroid inhalers at the first sign of asthma exacerbation is better than, and as safe as, continuing with the usual prescribed inhaled corticosteroid dose was updated. 33 From the eight studies included (1669 participants with mild to moderate asthma) the authors concluded that it is unlikely that increasing the dose of inhaled corticosteroid reduces the need for courses of systemic corticosteroids, hospitalisations or recovery time. The Cochrane review made it clear that its results were rated as being moderate to low quality, as the findings were uncertain across the studies and the studies conducted included very few participants in whom it could definitively be shown that increasing the dose was beneficial.

The benefit of an increase in inhaled corticosteroid treatment at the time of asthma control worsening is also supported by the benefit seen from studies evaluating Symbicort for maintenance and relief medication. Studies have shown that a variable dose of Symbicort, for use only by patients whose asthma becomes severe or whose asthma symptoms start to worsen, results in fewer acute exacerbations than a constant dose of maintenance treatment.

Strengths and limitations

This was an adequately powered RCT, with high follow-up rates for the primary outcome and a moderate adherence to the trial intervention.

The trial was open-label to research site staff and participants so, although the possibility cannot be ruled out that the treatment effect in the modified treatment group was enhanced, the inclusion of any placebo effect makes the study more relevant to real-life clinical practice.

The diary card completion of participants with at least one activation to zone 2 or above was not balanced between the two groups, with poorer completion noted in the usual-care self-management group than in the modified self-management group. This could have contributed towards a null bias, which could strengthen the positivity of the result. Overall, diary cards were not completed for the first activation to zone 2 by 34% of participants; this meant that information was unknown about adherence to the allocated self-management plans for 39% of participants in the usual-care group and 28% of participants in the modified self-management group.

A large number of missing data were missing for the secondary outcomes of area under peak flow curve and change in the Mini AQLQ score 2 weeks after activating zone 2 or above of the self-management plan. In addition, there was an imbalance in the number of participants with these outcomes available in the two allocated groups. Reassuringly, the results were consistent with the findings where more complete data were available, but these analyses still need to be interpreted with caution because of the potential risk of bias attributable to the missing data.

Finally, additional information on the acceptability of carrying an additional inhaler for the 80% of patients on a combination inhaler would have been useful information to have collected. Although no evidence was found that this led to early withdrawal from the study, it would be interesting to know how acceptable this would be for patients in real life as it could represent a barrier to the widespread use of such a self-management plan.

Generalisability

The study has good external validity as it was a pragmatic design that reflected normal clinical practice across both primary and secondary care in the UK. Participants used their existing asthma medication, and those in the modified self-management group either increased the dose of their corticosteroid inhaler or added an extra corticosteroid inhaler, depending on whether their usual treatment comprised corticosteroid or a combination of corticosteroid/long-acting beta agonist inhaler. It is important to note that throughout the trial the site staff ensured that the self-management plans were well explained and supported, and if this was not carried through into routine care the results may not apply.

Participants were recruited from 207 UK centres, both hospitals and GP surgeries, covering a range of urban and rural settings. This trial included people aged ≥ 16 years with chronic asthma who had had at least one acute exacerbation in the previous 12 months. The trial was inclusive of patients who were prescribed to the upper limit of the licensed dose of their maintenance steroid, so participants on very high doses of maintenance steroids were included in the trial. The study identified a group of patients at risk of further exacerbation and, therefore, with most to gain from following a self-management plan.

It is important to note that the exacerbation rate in the usual-care group was higher than expected, which suggests that the study recruited a sample of patients whose asthma was more severe than initially anticipated. So despite the percentage of participants on the modified self-management plan having an exacerbation being lower than that in the usual-care group, the overall exacerbation was still high (45%). As children were not included in this study it is not known if this intervention would be beneficial to treat asthma symptoms in children.

Main conclusions

This is the largest independent randomised controlled trial to assess the clinical effectiveness and cost-effectiveness of an asthma self-management plan that advises participants to increase fourfold their inhaled corticosteroid at the point at which asthma symptoms deteriorate. Both the clinical and economic analyses show this approach to asthma control to be effective for participants and health providers.

Implications for practice

The trial has shown that the use of an asthma self-management plan that advises patients to quadruple their dose of inhaled corticosteroid at the point of asthma deterioration is effective in reducing exacerbations that require unscheduled health-care consultations and the use of systemic corticosteroids in those patients who identify as having exacerbated within the last year, as well as proving cost-effective for health-care providers. Although quadrupling the corticosteroid dose appeared to be clinically effective and cost-effective across the licensed dose range, the systemic effects resulting from this advice in patients using high-dose inhaled corticosteroids need to be considered, and widespread adoption in these patients is not recommended. Clinical commissioners and national and international guideline developers can now be encouraged to make informed decisions regarding the use of self-management plans that advise a fourfold increase in inhaled corticosteroid dose on the basis of these robust findings. As only 15 patients need to be trained to use such a self-management plan to prevent one severe exacerbation, the study showed that all patients on low to medium doses of inhaled corticosteroids, especially if they have had an exacerbation in the last year, should be encouraged to follow such a plan.

Contributors to the FAST

Contributions of authors

Tricia McKeever (Medical Statistician) analysed the clinical effectiveness data and prepared the results for publication.

Kevin Mortimer (Senior Lecturer and Honorary Consultant in Respiratory Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Lucy Bradshaw (Medical Statistician) analysed the clinical effectiveness data and prepared the results for publication.

Rebecca Haydock (Trial Manager) managed the trial from November 2015 and drafted the final report.

Ian Pavord (Professor of Respiratory Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Bernard Higgins (Consultant Physician) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Samantha Walker (Deputy Chief Executive and Director of Research and Policy) contributed to the development of the self-management plans, trial protocol and important intellectual content of the report.

Andrew Wilson (Clinical Professor) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

David Price (Chair of Primary Care Respiratory Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Mike Thomas (Professor of Primary Care Research within Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Graham Devereux (Professor and Honorary Consultant in Respiratory Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Christopher Brightling (Senior Research Fellow and Clinical Professor in Respiratory Medicine) was a site principal investigator and contributed to the development of the grant application and trial protocol, and contributed important intellectual content to the report.

Charlotte Renwick (Health Economist) analysed the clinical cost-effectiveness data and prepared the results for publication.

Steve Parrott (Health Economist) contributed to the trial protocol, analysed the clinical cost-effectiveness data and prepared the results for publication.

Eleanor Mitchell (Senior Trial Manager) oversaw the trial management and contributed important intellectual content to the report.

Lelia Duley (Director Nottingham Clinical Trials Unit) contributed to the development of the grant application and trial protocol as well as important intellectual content to the report.

Tim Harrison (Professor and Honorary Consultant) was the chief investigator and co-authored the final report.

Publications

Skeggs A, McKeever T, Duley L, Mitchell E, Bradshaw L, Mortimer K, et al. Fourfold Asthma Study (FAST): a study protocol for a randomised controlled trial evaluating the clinical cost-effectiveness of temporarily quadrupling the dose of inhaled steroid to prevent asthma exacerbations. Trials 2016;17:499.

McKeever T, Mortimer K, Wilson A, Walker S, Brightling C, Skeggs A, et al. Quadrupling inhaled glucocorticoid dose to abort asthma exacerbations. NEJM 2018;378:902–10.

Data-sharing statement

All data requests should be submitted to the corresponding author for consideration. Access to anonymised data may be granted following review.

Patient data

This work uses data provided by patients and collected by the NHS as part of their care and support. Using patient data is vital to improve health and care for everyone. There is huge potential to make better use of information from people’s patient records, to understand more about disease, develop new treatments, monitor safety, and plan NHS services. Patient data should be kept safe and secure, to protect everyone’s privacy, and it’s important that there are safeguards to make sure that it is stored and used responsibly. Everyone should be able to find out about how patient data are used. #datasaveslives You can find out more about the background to this citation here: https://understandingpatientdata.org.uk/data-citation.

Disclaimers

This report presents independent research funded by the National Institute for Health Research (NIHR). The views and opinions expressed by authors in this publication are those of the authors and do not necessarily reflect those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care. If there are verbatim quotations included in this publication the views and opinions expressed by the interviewees are those of the interviewees and do not necessarily reflect those of the authors, those of the NHS, the NIHR, NETSCC, the HTA programme or the Department of Health and Social Care.

Wider societal costs

Only 41% of participants in both intervention groups reported their annual income and low numbers of participants reported costs of travel; therefore, an available case has been explored in this analysis to make the most of the available data. Table 36 shows the available cases for the number of days given up because of illness and the mean travel times to GP or hospital appointments as reported in the questionnaire. Using the participants’ number of GP visits and hospital outpatient visits, a mean total travel time per participant for each kind of appointment was calculated. There was little difference between intervention groups in the percentage of participants in paid employment (modified = 47% and usual care = 48% at 6 months). As for productivity loss, when looking at both time points together, there was little difference in the mean number of days given up as a result of respiratory illness between intervention groups.

Table 36 also shows how mean travel times to GP surgeries are mostly equal between groups at both time points; however, the modified self-management group has a slightly lower mean travel time per participant when the number of visits are taken into account, and both intervention groups have a lower travel time at 12 months. Once more, similar average travel times are seen for hospital outpatient visits and, when using the number of visits, there was little difference between the intervention groups and follow-ups.

The mode of travel to the GP surgery was mostly equal between intervention groups at both 6 months and 12 months, with car and walking the most popular methods. The majority of participants in both intervention groups and at both time points reported taking 0–15 minutes to reach the GP surgery [modified, 67% (6 months) and 64% (12 months); usual care, 61% (6 months) and 60% (12 months)]. Once again, the mode of transport to hospital outpatient visits was, on the whole, equally distributed between the two intervention groups, with both groups reporting more bus use than when travelling to the GP surgery. In both groups more participants reported having to travel for over 1 hour [modified, 15% (6 months) and 21% (12 months); usual care, 13% (6 months) and 17% (12 months)]. The percentages indicated the burden and time lost for travelling to hospital outpatient visits for both groups, as reflected in Table 36. Table 37 shows the mean travel costs per participant as reported and then how this translated into a total cost depending on their number of visits. The total travel costs are slightly greater for those in the modified self-management group when both time points are considered. This is because one participant in the modified self-management group reported very high costs for hospital outpatient visits.

A speculative estimate for the total cost at each follow-up time point has been calculated using productivity loss resulting from time off work and time spent travelling to appointments, as well as the travel costs reported in Table 37. However, the total numbers were very small because of the missing data. Table 38 shows the costs for the modified self-management group to be slightly higher than for the usual-care group at both time points. To make the most of the available data, the analysis was also repeated using the UK average wage rate (£31,800) to try and minimise the missing data caused by the poor report rate of income. 33 This resulted in a total of £220 (SD £1545, n = 487) for the usual-care group and £196 (SD £962, n = 504) for the modified self-management group.