Physiotherapy-led, community-based airway clearance services for people with chronic lung conditions: a retrospective descriptive evaluation of an existing model of care

In 2016 a metropolitan local health network, the Southern Adelaide Local Health Network (SALHN) identified a need to reduce the number of hospital admissions associated with exacerbations of chronic lung disease and to transition appropriate hospital-based outpatient services to community settings. One of the key objectives to meet this need was to develop a dedicated, physiotherapy-led, ACS in the community to assist people with chronic lung conditions to self-manage their cough and mucus hypersecretion and improve their health-related quality of life. The service plan was to provide most occasions of service on-site in a face-to-face community clinic. The capability to provide the service in a client’s home or via phone/telehealth platforms was available and encouraged when client frailty prevented clinic-based care.

Ethical approval for this retrospective evaluation was sought and approved by the Southern Adelaide Clinical Human Research Ethics Committee (HREC) (27/9/2022, ID:175.22) and the University of South Australia HREC (12/10/2022, ID:205,044).

SALHN provides care for more than 378,000 people [23]. One major tertiary institution provides hospital-based care for this region, around which rehabilitation and community service hubs are based. Prospectively, workforce planning for this ACS designated one full time equivalent (FTE) respiratory physiotherapist to operate a dedicated weekday clinic (at set days/times) from within these community hubs. The service aimed to provide education on the role of airway clearance in self-management, initiate airway clearance strategies/prescribe devices for suitable participants and address any associated respiratory symptoms through referral to relevant allied health disciplines and respiratory nursing services that were co-located on site. The ACS was publicly funded with no costs for service or devices to the consumer.

While the service and routine service data collection officially started on 26/6/16, for this retrospective audit, data was included from clients who attended for the first time during the first five complete calendar years of the service. All adults with a chronic lung condition referred to the ACS for the first time between 1/1/2017 and 31/12/2021 were eligible for inclusion within this evaluation. Repeat referrals for these included individuals and any associated data were excluded from analysis. Referrals were accepted from public and private physicians, nurses and allied health professionals and were grouped into referral source categories (Fig. 1). Triage categories of referral urgency were pre-determined (1 = 0–30 days, 2 = 30–60 days, 3 = 60 + days) and were based on the information provided on the referral and/or the electronic medical record. Decisions about which data would be collected and how were made as part of the planning phase of the service. Patient-reported outcome measures (PROMs) were prospectively chosen by the service key stakeholder group (including health professionals from physiotherapy, nursing, medical and general practice disciplines) and in accordance with other local integrated services (PR, GPs) to reduce participant burden and duplication. The choice of PROMs was based on suitability and validity in older people with chronic lung conditions anticipated to be seen by the service [24‐27], and the availability of the PROMs free of charge. Service-specific forms (registration and satisfaction surveys) were prepared before the service was implemented. While this ACS was open to referrals for adults with any chronic lung condition, people with CF and motor-neuron disease were not referred as they were managed by specialised state-wide disease specific services.

Data collected for each person eligible for inclusion in this audit was entered on to a spreadsheet by a single administrator (Table 1). Referral and participant demographic data were recorded from the paper referral and the electronic medical record. The individual’s diagnosis or condition and reasons for referral were transcribed verbatim from the referral. Individuals were posted registration forms and PROMs appropriate to their specific health condition. All referred individuals received a Leicester Cough Questionnaire (LCQ), a Depression, Stress and Anxiety Scale-21 (DASS21) and an ACS registration form. The ACS registration form was purpose designed for participants to self-report symptoms (sputum quantity, colour, viscosity) and associated co-morbidities (reflux, sinusitis) and bring to their first appointment. Those with COPD and/or bronchiectasis received the COPD Assessment Tool (CAT) and/or the Quality of Life – Bronchiectasis (QOL-B) questionnaire respectively.

Care was individualised and there was no predetermined limit to the number of sessions. Where possible, goals of care were agreed between the treating physiotherapist and the participant at the first appointment. The choice of airway clearance technique, device prescription and/or lifestyle modification strategies employed were based on clinical need/effectiveness, adherence to the therapy, and client preference. Symptoms such as incontinence, musculoskeletal pain and reflux were addressed on an individual basis with on-referral to other health professionals as required. Participants attended the program until the agreed goals of care were achieved. Upon formal discharge from the service, follow up PROMs (for participants completing them prior to the service) and a satisfaction survey were sent by post with a reply-paid envelope for return (Supplementary Data, S1). The satisfaction survey included six statements (using a 5-point Likert scale) and free text feedback options inviting people to report on the service (Fig. 2). Participants voluntarily completed PROMs at home and returned them to the site (at initial face-to-face appointment) or via post (after discharge). The treating clinician scored the PROMs and included the results in the medical documentation as appropriate. The treating physiotherapist produced a discharge summary for the participant’s general practitioner (GP) and referrer.

The variables of interest in this descriptive analysis were grouped into domains: referrals (source, date); participants (demographic characteristics); service provision (wait times, occasions of service, modes of delivery); outcomes (PROMS, satisfaction) (Table 1). Measures of socioeconomic status (Index of Relative Socio-Economic Advantage and Disadvantage (IRSAD) [28] and accessibility/remoteness of participants (Accessibility/Remoteness Index of Australia (ARIA) [29] were derived from participants’ residential postcodes. The frequency and nature of missing data was reviewed: where possible missing hospital system data were retrieved from alternate platforms. PROMs and satisfaction surveys relied on voluntary participant completion and were not routinely checked for missing items or discussed with the participants (Table 1). PROM scores were calculated as recommended by the instrument developers [24‐27]. No imputation was used for missing data.

The sampling frame considered all persons referred to the ACS and was subdivided into first-time referrals (included in evaluation analysis) and repeat referrals (excluded from all subsequent analysis). Differences (in demographics, referral source and referral-stated condition) between those referred who did and did not participate in the service were explored using chi squared tests (α = 0.05). For people who participated in the ACS (i.e., attended at least one appointment) descriptive statistics (frequencies, percentages, central tendency and spread) were used to report all categorical and continuous data on demographics, ACS service provision and self-reported symptom information.

The response rate of completed pre- and post- service PROMs was reported as a percentage of the total number of participants sent them by post. Where available (i.e., questionnaires voluntarily returned via reply-paid envelopes), paired (pre-post) PROM data for individuals was explored for clinical significance according to the minimal clinically important difference (MCID) reported for each instrument [24‐27]. Pre-post changes for individuals reaching the MCID for the LCQ (+ 1.3 point change [26]), DASS (-5 point change [25]), CAT (-2 point change [24]) and QOL-B (respiratory domain ≥ 8.0 points [27]) were interpreted to indicate an improvement in clinical condition. The number of individuals reaching the MCID for each PROM was reported and expressed as a proportion of the number of the paired PROM questionnaires returned. The satisfaction survey response rate was reported as a percentage of the total number of participants who were discharged from the service and survey responses were reported descriptively.

More women (n = 792/1335, 59%) than men (n = 543/1335, 41%) were referred to the ACS with the average age of referees being 69 years. Most resided in major cities (n = 1114/1335, 83%) and inner regional areas (n = 177/1335, 13%) with near equal distribution across the IRSAD quintiles 2 (disadvantaged) to 5 (most advantaged). There were no differences between those referred who did and did not attend the service in age, gender, geographic location, socio-economic disadvantage, or condition/reason for referral. Referral source was different in people who did or did not attend the service (χ2 = 22.563, p = 0.0004) with those referred from public SALHN inpatient wards less likely to attend (21 of 69 referrals from this source did not attend, 30%). The three most common ‘reasons’ for referral to the ACS were bronchiectasis (n = 649/1335, 49%), COPD (n = 408/1335, 31%) and asthma (n = 247/1335, 19%).

Self-reported symptom information obtained from the ACS registration forms (returned by 789/1157, 68%) showed that daily mucus load was common (n = 500/789, 63%), and not only associated with infections (n = 468/789, 59%). Mucus was described as thick (n = 239/789, 30%), sticky (n = 226/789, 29%) and difficult to expectorate (n = 456/789, 58%). Symptoms of reflux or heartburn (n = 407/789, 52%), sinusitis or rhinitis (n = 223/789, 28%), leaking urine (n = 267/789, 34%) and musculoskeletal pain (n = 223/789, 28%) on coughing were commonly reported.

A total of 2996 occasions of service (face-to-face clinic n = 2108 (70%), phone n = 736 (25%), telehealth n = 99 (3%), home visit n = 53 (2%)) were delivered to the 1157 patients attending the ACS. On average the initial appointment occurred within 36 calendar days of being referred; people attended three times and were discharged within 146 calendar days. A discharge summary to the GP and referrer was completed for most people (n = 869/1155, 75%).

The three most common airway clearance devices that were prescribed and provided were oscillating positive expiratory pressure (PEP) devices; the Aerobika (n = 525/1157, 45%), bubble PEP (n = 263/1157, 23%) and the Acapella (n = 127/1157, 11%). The services most on-referred to were PR (n = 116), speech pathology (n = 66), dietetics (n = 10) and psychology (n = 9).

Individuals completed both pre and post impact questionnaires around a third of the time (CAT 33%; QOL-B 35%; DASS-21 22%; LCQ 24%) with at least half of responders reporting an improvement in respiratory symptom outcomes during this time that was consistent with the MCID (CAT 53%; DASS 21 27%; LCQ 50%; QOL-B respiratory domain 61%) (Table 4). There were no observable differences (gender, age, referral source and number of appointments with the service) between those completing PROMS and those that did not (Supplementary Data, Table S7). Satisfaction surveys were returned by 289 individual ACS attendees who were formally discharged from the service (289/1155, 25%). Most participants (> 95%) who responded indicated they agreed (agree or strongly agreed) with each statement; less than 2% indicated neutral responses or disagreement (disagree, strongly disagree) (Fig. 2 and Supplementary Data, Table S3).

Data collection reflected the practices of the service, not a clinical trial, and assumed that planned administrative processes were adhered to. Planned data analysis for PROMs focussed upon the proportion of participants completing both pre and post outcomes assessments, and meeting or exceeding established MCIDs for specific instruments. Testing for statistical differences pre to post service usage was not planned as completion rates of PROMs was voluntary, completion rates were likely to be less than 80% of possible participants, and a high potential for missing data and/or PROMs. The observed response rate for people submitting PROMS both pre and post ACS (paired data) as a percentage of all pre-ACS PROMS was less than 40% across specific PROMS. As the service plan for data collection and establishment of a database were for the purposes of recording and tracking service provision (e.g., referral source, diagnosis, dates of appointments, budget outlay on devices), not all aspects of clinical intervention delivery or outcomes were recorded. For example, data on the frequency of other common airway clearance interventions such as general education, the Active Cycle of Breathing Technique [41] and forced expiratory techniques were not collected by the ACS, nor reported on in this service evaluation. Similarly, first line education about management of the clients’ self-reported concurrent symptoms (e.g., reflux, heartburn, sinusitis, pain or urine leakage) was not captured in the audit data. It is not possible to interpret service efficacy from pre-service symptom questionnaires, pre-post service PROMs, and post-service satisfaction, as completion was voluntary and there may be unknown differences between those who participated or did not. Diagnosis and referral reason were recorded from the referral and were not confirmed which may not reflect all known respiratory co-morbidities in the cohort. Consumers were not consulted in the early-stage planning phases of this ACS. An analysis of the impact of this service on consumer healthcare usage would provide an important economic viewpoint, that coupled with consumer and healthcare professional insights into the ideal model of service delivery may see this service evolve in the future.