Development of the Lymphoedema Patient Reported Outcome Measure (LYMPROM)

Chronic oedema due to an insufficiency of the lymphatic system is known as lymphoedema (Moffatt et al, 2003). Regardless of aetiology, the lifelong effects of lymphoedema can lead to significant physical, psychological and social difficulties (Greene and Meskell, 2017; Thomas et al, 2020a), such as pain, cellulitis, lymphorrhoea and impaired quality of life (QoL) (Tobin et al, 1993; McWayne and Heiney, 2005; Stolldorf et al, 2016).

Lymphoedema represents a significant challenge for patients, healthcare providers and services; and the magnitude of this problem may not be fully recognised (Moffatt et al, 2003; 2017; Thomas and Morgan, 2017). Risk reduction, early intervention and ongoing management of lymphoedema are costly and labour intensive (Morgan et al, 2005; Ridner et al, 2012; Todd, 2013). The four cornerstones of management—skin care, exercise, massage and compression (Lymphoedema Framework, 2006)—rely on effective partnership, motivation and engagement (Bogan et al, 2007).

Patients with lymphoedema require lifelong lymphoedema care. Admissions due to lymphoedema cost English health services more than £178 million (National Lymphoedema Partnership, 2019), with the annual costs of managing wounds (£4.5-5.1 billion (Guest et al, 2017)) and cellulitis (£172-254 million, (Atkin, 2016)) compounding estimates and amplifying the burden on patients with lymphoedema. Strategies to understand service provision and improve education and management are under way (Thomas et al, 2017; 2020b). These initiatives should reduce the financial burden, increase efficiencies and improve patients' quality of life (Thomas et al, 2017). A recent audit of data in Wales suggested that an education programme could save over £20 000 in management costs per patient based on, for example, reduced district nurse visits and cellulitis episodes.

Patient-reported outcome measures (PROMs) support patients to reflect on their condition, which helps therapists to identify patient priorities, personalise holistic care (Weiss and Stray, 2002; Mangin et al, 2016) and drive ‘value-based health care’ (Porter, 2009; Value in Health Programme, 2021). With escalating patient referrals and a growing ongoing caseload the demand–capacity gap concerns remain problematic. A potential solution to capacity may be to focus care on the people whose needs are the highest by utilising PROMs. Thus, those patients with low-scoring PROMs (low impact of lymphoedema on day-to-day living) could be discharged, with open access to return if need be. This would allow services to focus on the care needed for those patients who scored high on PROMs (high impact of lymphoedema on day-to-day living). This would encompass value-based health care: seeing the patients whose needs are the greatest. This article describes the development of the Lymphoedema Patient Reported Outcome Measure (LYMPROM).

Prioritising PROMs

Historically, PROMs were collected by lymphoedema services using the All-Wales Lymphoedema and Lymphatic Venous Anastomosis Assessment and Review forms. The pool of relevant PROMs was first identified from the literature suggesting objective and subjective items impacting on social and emotional QoL. This pool of items was reviewed and refined with clinical experts (18) and stakeholders (7) during Lymphoedema Network Wales benefit workshops in 2014.

Based on clinical need, most lymphoedema patients are reviewed twice a year during a 45-minute consultation. To avoid burdening the patients the panel of experts felt that the PROM should be ‘easy’ to complete. With consensus, the PROM included 12 patient priorities/items covering three domains:

• Physical health: pain; heaviness of limb/s, life, home life, work, shopping for clothes/shoes, mobility
• Social health: hobbies, holidays
• Emotional health: anxiety, body image, sexuality/desirability/intimacy.

The impact of each item was rated using an 11-point distress thermometer with anchored numbers, ranging from 0 (no impact) to 10 (extreme impact). To ensure all patients had an opportunity to report any other concerns a free-text box was also included.

Existing general QoL forms and lymphoedema PROMs do not capture all 12 of these patient priorities (Table 1). Furthermore, most PROMs were not fully validated or were validated for use with a specific aetiology or area of lymphoedema; others were lengthy.

Although existing QoL forms capture the impact of lymphoedema, they do not support services in improving the demand–capacity gap by focusing on value-based health care. As the Lymphoedema Network Wales demand–capacity gap increased, issuing PROMs became fundamental to the planning and delivery of services, ensuring that the patients with the greatest need were prioritised. Therefore, the 12 patient priorities were extracted from clinic-based paperwork into the specific LYMPROM document. This was designed to be completed by the patient at home and returned to services, to explore whether a face-to-face appointment was necessary based on their PROM score.

To ensure that LYMPROM was fit for purpose an evaluation was undertaken, considering its feasibility and acceptability along with initial estimates of its validity and reliability.

Developing and evaluating LYMPROM

This project was reviewed at the Research and Development Joint Study Review Committee at Swansea Bay University Health Board and ratified as a service evaluation (14/04/2020).

An iterative approach was taken:

• Feasibility and acceptability of LYMPROM:
• Internal consistency
• Floor and ceiling effects
• Expert review of LYMPROM:
• A workshop with experts in lymphoedema and value-based health care plus a focus group with patient representatives
• Face, comprehension and content validity.

Data collection

During a 3-month period in 2019, a lymphoedema practitioner from each of the seven Welsh NHS Health Boards was asked to collate completed LYMPROM from around 20 of their adult lymphoedema patients attending for an appointment. LYMPROM was independently completed on paper by patients and returned either by post or in the clinic waiting–room.

Adult patients with primary or secondary lymphoedema attending for new and follow-up appointments completed LYMPROM as per their usual care, enabling clinical knowledge to combine with patient conversations to personalise the care received. These patients were also asked if their anonymous LYMPROM data could be shared for this evaluation. Non-participation did not affect usual care.

Anonymised copies of LYMPROM were scanned and sent to the national lymphoedema research specialist, to be reviewed and data entered into Excel. Data were imported into Statistical Package for Social Sciences (SPSSv26) for statistical analyses. Lymphoedema outcome data or patient demographic data were not collected.

Feasibility, acceptability and data completeness

In total, 128 anonymous LYMPROM forms were reviewed from adult lymphoedema patients, with 66% of patients (84/128) completing all 12 items on LYMPROM.

Feasibility of LYMPROM was established; no patient refused to complete it as part of their usual care and all patients permitted use of their anonymous data for evaluation. The majority of therapists asked patients to independently complete LYMPROM while in the waiting room. Three items on LYMPROM (work, hobbies and sexuality/desirability/intimacy) were left blank by more than 8% of patients and four items were annotated as non-applicable by more than 2% of respondents (Table 2). One patient annotated that ‘sexuality/intimacy/desirability’ was too personal to answer. A minority of items answered could not be coded, for example where a range was given or a selection was made outside of a box. Less than 2% of participants entered a range of scores (on pain, heaviness and sexuality).

Supplementary information was provided by 9% (12/128) of patients, including problems of itchiness (1), loss of appetite (1) or loss of sleep (4). For two of the patients, the sleep problem was due to pain already reported on LYMPROM. Elsewhere, patients used the free-text box to explain scores such as taking longer to do activities or challenges on aspects of social living.

A minority (5/128) of patients reported no burden of lymphoedema as scored 0/10 in all categories, while 27 patients signposted a problem (scored at least 1/10) in all 12 items, of which over a quarter (7/27) scored at least 5/10 in all 12 items. No one scored 10/10 in all 12 categories. The average number of problems per patient was 7.85/12 (items were scored at least 1/10). Three of the most burdensome problems reported by patients were shopping for clothes or shoes (average score 4.9/10, n=120); body image (4.2/10, n=121) and mobility (4.0/10, n=121).

Internal consistency

Internal consistency of LYMPROM was assumed if Cronbach's alpha was >0.7 (Cortina, 1993; Peterson, 1994). With missing data excluded, the overall Cronbach's alpha score was excellent at 0.958 (n=84) suggesting strong correlation (12 items), with no negative correlations observed (Table 3). Owing to the small number of items, it was not meaningful to pool domain items, therefore health-related QoL was considered as the underlying construct for all items. All inter-item correlations were above 0.4. For reference 0–0.4 is a mild correlation, 0.4–0.7 is a moderate correlation, 0.7 and above is a strong correlation. Four inter-item correlations were marginally in excess of 0.8: life and home life, 0.831; anxiety and home life, 0.854; anxiety and life, 0.829; mobility and shopping, 0.836. However, no items were identified for removal.

Floor and ceiling effects

Looking at overall LYMPROM scores, no floor or ceiling effects were identified: 4% (5/128) of patients reported the lowest possible score (no burden associated with lymphoedema 0/120) and no patient scored the highest possible score (maximum burden with lymphoedema 120/120). Although a range of scores were received for each item, floor effects are identified at the item level: more than 15% reported the lowest score for 11/12 items (Terwee et al, 2007; Lim et al, 2015). Ceiling effects (>15%) were identified in 2/12 LYMPROM items (Table 4).

Expert review of LYMPROM

To evaluate LYMPROM further, a workshop with 16 experts (14 lymphoedema health professionals and two value-based healthcare staff) and a focus group with four patient representatives (with lymphoedema affecting any area of their body) identified two important outcomes (personal care and finances) not originally captured by LYMPROM.

Following suggestions from the workshop and focus group, and with analysis of the data gathered, further adaptations were made to aid patients' understanding of LYMPROM, including:

• Detailed patient instructions, as well as item definitions and examples
• The ‘effect’ of lymphoedema was altered to ‘impact’
• ‘Mobility’ was re-defined as ‘walking’
• Patients were asked to reflect on the past 4 weeks when answering each question
• A visual aid for rating was added
• The ordering of questions was reviewed and potentially sensitive items were moved towards the end of LYMPROM
• Sexuality was removed, but intimacy/desirability retained
• Non-applicable options were added to relevant questions and placed away from the rating scale to minimise response error
• As ‘life’ and ‘home life’ were highly correlated, ‘life’ was removed with ‘home life’ remaining to avoid overlap, giving rise to the 13-item LYMPROM.

Examining face and content validity

Nine lymphoedema experts and five patient representatives with lymphoedema from across Wales reviewed the 13-item LYMPROM using the content validity index (CVI) with a four-point scale of agreement (Lynn, 1986; Polit and Beck, 2006; Gabe and Jordan, 2014) (Table 5).

Overall, items within LYMPROM appear to have face and content validity and it appears to adequately sample patient priorities. However, there may be some overlapping questions. The majority of items reached an item CVI ( ) in excess of 0.7 with an overall scale CVI/Average (s-CVI/Ave) of 0.94 (Polit et al, 2017).

The majority of respondents (11/14) added further comments covering benefits of LYMPROM and suggested revisions. Based on the content and face validity the term ‘discomfort’ was added to the ‘pain’ category and the ordering of the items was amended to group ‘hobbies’ and ‘holidays’ together, along with making ‘home life’ the first category. As a result of the removal and addition of items, the final version of LYMPROM included 13 items over three domains (physical, social and emotional).

Based on all the evaluations, the final version of LYMPROM is now ready for a larger study, further testing reliability and validity along with digital integration.

Discussion

The aim of this article was to describe the development of LYMPROM as used within routine lymphoedema care. LYMPROM addresses the gap in existing lymphoedema PROMs based on reported patient priorities. Patients provided invaluable insight on the outcomes of interest (Wiering et al, 2017), along with their relevance and clarity. Routine use of LYMPROM as a clinical tool was acceptable and feasible, supporting patients to report their ‘lymphoedema burden’ and the therapist to plan care:

‘[It] … enables the practitioner to assess the patient holistically. It may highlight areas of concern that would have not been discussed previously and may highlight areas the patient may be too embarrassed to address.’

ID08

Missing LYMPROM items do not appear to be due to lack of acceptability, willingness to complete or difficulty with the question(s). Annotations added by many patients suggested that some items were not relevant or that the difficulty lay in selecting the appropriate response options. For example, impact on employment was considered non-applicable for some, yet voluntary work or caring commitments were not considered. Accordingly, all questions were supplemented with an explanation and a non-applicable response choice was added to problematic items. ‘Sexuality/intimacy/desirability’ was also refined to ‘intimacy/desirability’ because the three terms could lead to confusion. Less than 10% of patients added information in the free-text box, which may be testament to the adequate coverage of patient-reported measures within LYMPROM. However, the CVI (Table 5) suggested that there may be overlapping questions. Free-text comments were reviewed to target resolution. However, no item was earmarked for removal and only one participant indicated an ‘overlap’ between holidays and body image. Another explanation for this low score item-CVI may be a switch in layout within the CVI form: positive views were switched from the right-hand to the left-hand column. Two further items suggested for inclusion by experts covered a new domain on experience of prescribed compression garments, while another explored the role of comorbidities on reported burden. These were not added because it was considered a separate project and will be explored using audit.

Ensuring LYMPROM is comprehensive yet succinct and easy to use is a challenge. Ideally patients should complete LYMPROM before attending their lymphoedema appointment, allowing therapists to prioritise face-to-face appointments based on patient need. However, this was not feasible because the majority of lymphoedema services incorporated LYMPROM during clinic time. Positively, work exploring opportunities to digitise the tool and enable paper-light completion at home by patients is under way, and will be reported at a later date. Overall, internal consistency was high, however floor effects at the item level are a concern, which may hinder reliability and represents a concern to content validity. Indicating that the scale may not be sufficiently comprehensive and unable to detect change or distinguish between patients with the lowest score. However, these findings may reflect the considerable volume of patients with mild lymphoedema being somewhat asymptomatic. The floor effect also highlights the potential benefit of using LYMPROM to increase capacity within lymphoedema services because these patients should be discharged. Anecdotally, therapists felt more confident to discharge patients when they reviewed the low LYMPROM scores.

The clinician–patient relationship is an important component of shared decision-making and patient involvement (Department of Health, 2010). It is beyond the scope of this article to explore the fullness of impact of LYMPROM, but we anticipate that patient-led PROMs may impact on patient empowerment, engagement, quality (Nelson et al, 2015; Greenhalgh et al, 2018) and value-based health care. Alongside an in-depth evaluation of LYMPROM based on the COSMIN initiative (COnsensus-based Standards for the selection of health Measurement INstruments) (Mokkink et al, 2010; 2019), work is also under way to explore long-term impact and patient experience; along with scoring LYMPROM to inform clinical decision-making.

Limitations

This work was undertaken during 2020, including the period covered by the COVID-19 pandemic. As a result, some of the planned evaluations were deferred. Although this evaluation describes some of the initial testing of LYMPROM, ongoing evaluations will consolidate findings with COSMIN guidelines. It is not possible to say that the priorities identified using LYMPROM are entirely due to lymphoedema: comorbidities, and socioeconomic and cultural aspects cannot be ruled out. However, the aetiology of a problem is likely to be of secondary importance to patients: the crucial issue is prioritised personal care. Furthermore, this evaluation reviewed cross-sectional data for adult patients only and did not collect information on the area or aetiology of lymphoedema. Thus, it is difficult to understand whether the findings apply to different groups of patients with lymphoedema.

Conclusion

LYMPROM promotes quality in care, epitomises value-based health care and collates patient priorities to focus need and reduce waste. Working with patients, important patient outcomes have been captured in a succinct tool. Assessing the burden of lymphoedema by the ‘size’ of the swelling alone may be problematic. This initial review suggests that everyday challenges such as shopping for clothes or shoes, body image and mobility should also be considered when planning patient care. LYMPROM is easy to complete and most people should need no more than 10 minutes to do so. More research is under way to fully understand the validity and reliability of LYMPROM, along with its translation into Welsh and its integration on a digital platform.

KEY POINTS

• Patient-reported outcome measures can be used to plan value-based health care
• LYMPROM (Lymphoedema Patient Reported Outcome Measure) is based on patient need and its use in lymphoedema care is acceptable and feasible
• LYMPROM helps patients to prioritise goals and supports therapists to plan appropriate, timely care
• Evaluation of the validity and reliability of LYMPROM is ongoing

CPD reflective questions

• Consider the benefits of using PROMs specific for lymphoedema
• How does using PROMs support the delivery of value-based health care?
• What are the challenges of using PROMs with patients?