The COVID‐19 Yorkshire Rehabilitation Scale (C19‐YRS): Application and psychometric analysis in a post‐COVID‐19 syndrome cohort

Abstract As our understanding of the nature and prevalence of post‐coronavirus disease 2019 (COVID‐19) syndrome (PCS) is increasing, a measure of the impact of COVID‐19 could provide valuable insights into patients' perceptions in clinical trials and epidemiological studies as well as routine clinical practice. To evaluate the clinical usefulness and psychometric properties of the COVID‐19 Yorkshire Rehabilitation Scale (C19‐YRS) in patients with PCS, a prospective, observational study of 187 consecutive patients attending a post‐COVID‐19 rehabilitation clinic was conducted. The C19‐YRS was used to record patients' symptoms, functioning, and disability. A global health question was used to measure the overall impact of PCS on health. Classical psychometric methods (data quality, scaling assumptions, targeting, reliability, and validity) were used to assess the C19‐YRS. For the total group, missing data were low, scaling and targeting assumptions were satisfied, and internal consistency was high (Cronbach's α = 0.891). Relationships between the overall perception of health and patients' reports of symptoms, functioning, and disability demonstrated good concordance. This is the first study to examine the psychometric properties of an outcome measure in patients with PCS. In this sample of patients, the C19‐YRS was clinically useful and satisfied standard psychometric criteria, providing preliminary evidence of its suitability as a measure of PCS.


| INTRODUCTION
The medium and long-term problems experienced by survivors of coronavirus disease 2019  are emerging, with over one million people in the UK who contracted COVID-19 reporting symptoms and functional problems more than 4 weeks after onset of the acute illness. 1 Almost 700 000 people report ongoing impact on their health and functioning more than 12 weeks after the acute infection, and this combination of symptoms and functional difficulties is recognized as a new syndrome called "Long COVID" (LC) or post-COVID-19 syndrome (PCS). 1 The most common symptoms include fatigue, breathlessness, pain, anxiety, and cognitive impairment, but there are over 200 reported symptoms affecting 10 organ systems. 2 One study following 143 individuals seven weeks postdischarge found 53% of patients reported fatigue, 43% breathlessness, and 27% joint pain. 2 A substantial number of people report limitations with their activities of daily living, with almost 130 000 patients stating that these limitations are severe. 3 Given the novelty and uniqueness of the syndrome, it is unsurprising that standardized assessments of functioning, disability, and health are lacking. While generic assessments are available, these have been shown to lack responsiveness or be useful only for discriminative purposes. 4,5 The long-term symptoms of COVID-19 might be predicted from infection is similar. 7 Five studies in our metanalysis included the Medical Outcomes Trust Short-Form 36-item generic health outcome measure (SF-36) as part of a basket of measures to try to determine health outcomes after the acute infection. The breadth of long-term symptoms in patients affected by the previous SARS outbreaks, and the emerging evidence of the long-term impact of COVID-19, means that a single, generic health outcome measure, or indeed a basket of measures, will be adequate to capture the breadth of these symptoms in a succinct way that is acceptable to patients and clinicians.
Furthermore, the responsiveness of the SF-36 in detecting clinical change when it occurred was small, limiting the utility of this measure in measuring the effect of an intervention.
Using symptoms and functional difficulties of PCS that were being reported by survivors of acute COVID-19 infection and the healthcare professionals involved in their care from across the clinical sites in the Yorkshire region, 7 we developed a condition-specific measure for PCS.
The COVID-19 Yorkshire Rehabilitation Scale (C19-YRS) is a 22-item patient-reported outcome measure designed to evaluate the long-term impact of COVID-19 across the domains of Activities and Participation of the International Classification of Functioning, Disability, and Health and evaluate the impact of PCS rehabilitation. 8 The C19-YRS now includes clinician-completed, self-report, and digital versions. 3 Content validity of the C19-YRS has been demonstrated, 7 and the C19-YRS is now used in the UK's first specialist PCS community rehabilitation service and 26 other National Health Service (NHS) PCS services in the UK. 9 This article describes the first stage in establishing the initial psychometric properties of the C19-YRS as an outcome measure for PCS using classical test theory. 10 Its ongoing development will investigate and address any problems with its psychometric properties using Item Response Theory (specifically, Rasch analysis). 11 This will explore the presence of differential item functioning, local dependency of items, and will examine unidimensionality and the YRS' responsiveness.

| METHODS AND DESIGN
This study was a prospective, observational study, and psychometric analysis of data captured from long COVID patients using the C19-YRS questionnaire. Long COVID patients were recruited from a community-  • Ability to complete a self-report C19-YRS as part of initial triage.

| Participants and recruitment
Although literacy and language ability were not initially screened unless highlighted by the referrer, support to complete the C19-YRS form was provided where necessary by a family member or carer, clinician, researcher, or a proprietary translation service used by the clinical service.
• Willing and able to consent for data to be used anonymously for research and/or service evaluation purposes. Consent was gathered via the first page of the C19-YRS form and did not affect a patient's access to treatment. Data are already collected in the service as part of routine initial evaluation to form a functional baseline and are documented in the patient's electronic clinical notes. anonymized, but details such as sex and age (not the date of birth) were included. The full process is illustrated in Figure 1. Anonymity was ensured by linking patient identifiable details to ID numbers on one Excel sheet, and full C19-YRS data sets on another. Only the ID numbers linked the two documents, which were removed before statistical team input.

| Exclusion criteria
It was not felt by the research team that formal training was required before contacting patients (to organize the provision of the C19-YRS) nor completing the measure via the phone, as it is possible to read questions verbatim from the C19-YRS. A standard operating procedure and telephone script were agreed upon, including allowing questions to be repeated for clarification with prompts provided for scale grading.

| Data quality
Data quality concerns the extent to which a scale can be administered successfully in the target sample. The C19-YRS data were examined for percentage missing items and the percentage of the sample for whom total scores could be calculated. For responders with missing items, imputed scores were not used. 12

| Scaling assumptions
Tests of scaling assumptions examine whether it is legitimate to sum item scores to generate scale scores. In order for a set of items to be legitimately summed to form a total score, a series of criteria should be satisfied. [13][14][15] We tested the C19-YRS against these criteria, which are: Items should be roughly parallel, that is, measure at the same point on the scale and have similar variance, otherwise they do not contribute equally to the variance of the total score. 16 A set of items is considered parallel when their item response option frequency distributions and their item mean scores and standard deviations are roughly similar. 14 Items should measure the same underlying construct, otherwise, it is not appropriate to combine them to generate a total score. 11 A set of items is considered to be measuring the same construct when each item's corrected item-total correlation, which is the correlation between each item and the total score computed from the remaining items in that scale, exceeds 0.30. 16 Items in the scale should contain a similar proportion of information concerning the construct being measured. This criterion is considered satisfied if the corrected item-total correlations exceed 0.30. 17

| Targeting
Targeting refers to the match between the distribution of health problems in the sample and the range of health problems measured by the scale. The better this match, the greater the potential for precise measurement. Targeting was evaluated by examining floor and ceiling effects, score distributions, and skewness statistics. Floor effects are the percentage of patients scoring 10 (most severe impact of symptom) and ceiling effects are the percentage of patients scoring zero (symptom not present). It is recommended that floor and ceiling effects should be less than 20% each on each item. 18

| Reliability
Reliability describes the extent to which scale scores are free from random error. Scales should generate reliable estimates of the construct being measured (internal consistency). Cronbach's α coefficient was used to determine this criterion. 19 Although a range of minimum values has been suggested, it is widely accepted that Cronbach's α should exceed 0.80 for group comparison studies. 15 3 | RESULTS

| Sample characteristics
Data for the analyses were obtained from 188 consecutive assessments of PCS patients. Patient details are given in Table 1. One patient was removed from the analyses because a significant number of answers were missing, presumed to be an oversight of the respondent, or due to symptoms impacting on the ability to complete the scale in full.
Follow-up contact was not made with patients unless an answer to the "overall health" question was suspected to be erroneously scored Patients' scores on the C19-YRS sub-scales are presented in Table 2. Fatigue was the most common complaint, with 97.3% of patients reporting fatigue of varying severity, followed by the onset of pain, which was not present before COVID-19 was contracted (94.3%). The most common new pain was muscle pain, which affected 70% of patients, followed by headache (67%), chest (64%), and joint pain (59%). Approximately one-third of patients also experienced new pain in their abdominal or other regions. Mental health problems were reported by 41% of patients, with 17% of these patients reporting respiratory or cardiac comorbidity. Respiratory or cardiac health issues, or both, were reported by 37% of patients. Swallowing, incontinence, skin rash, and fever were troublesome for very few the respondents.

| Data quality
Missing data for items were low (range 0.5%-19.8%). Subscale scores could be calculated for 67% of patients reporting symptom severity, 82% of patients reporting functional disability, 83% of patients reporting additional symptoms, and 98% of patients reporting overall health. Details of scores are given in Table 2.

| Scaling assumptions
Item response option frequency distributions were symmetric. Item means and standard deviations were similar indicating that they were roughly parallel (Table 3), although there was a greater range in symptom severity. Corrected item-total correlations exceeded 0.30 for all items except swallowing (0.24), incontinence (0.28), and skin rash (0.14), indicating that scaling assumptions were met for most items, including fever (0.33).

| Targeting
Scores spanned the range of the scale on admission and discharge and demonstrated good variability (Table 3). Results for some items demonstrated notable floor effects, especially for swallowing (72.7%), skin rash (66.8%), and fever (64.7%). There were no ceiling effects in any subscale. *Overall health was reversed scored compared to item severity, so that an overall health score of "10" reflected the best possible health, in contrast to item severity where "10" reflected the worst possible severity of the symptom.

| Reliability
nonhospitalized patients, to determine the stability of the psychometric properties and its potential as a measure of PCS. In this first round of preliminary testing, our results provide evidence for that potential. In the group studied, evidence was found for data quality, scaling assumptions, targeting, and reliability. The findings from this study provide useful information and illustrate the potential of the Despite these limitations, we are confident that the C19-YRS will turn out to be a useful addition to current assessments of post-COVID-19 in clinical studies, and could be used to complement clinician-rated measures of symptoms. Furthermore, the items in the scale provide qualitative information to clinicians to assist in targeting their clinical interventions to individuals' needs. It has advantages over other approaches, as it may be used in any setting, does not require an external rater, and is not laboratory-based or require special equipment. Most importantly it measures patients' perspectives. 27

| Further research
In future validations, as cases accumulate, the researcher will seek outpatients whose circumstances and perspectives provide a contrast to those already included to achieve maximum variety in clinical, social, ethnic, and personal circumstances and health/digital literacy.
Subsequent psychometric testing will use Rasch analysis to determine whether the scale meets the fundamental axioms that define scientific measurement and permit the transformation of raw (ordinal) scores to interval level measurement. 11 Further evaluations will examine the short-and long-term responsiveness of the scale to changes in symptom severity and the overall impact of rehabilitation on PCS. This will also determine the minimal clinical important difference of the scale that correlates to clinical improvement or deterioration of the condition reported by patients.

| CONCLUSION
This is the first study to examine the psychometric properties of a PCS-specific outcome measure that captures and evaluates the symptoms experienced by patients. In this sample of patients, the C19-YRS was clinically useful and satisfied standard psychometric criteria. The C19-YRS shows good internal consistency, and scaling and targeting assumptions were satisfied. This provides preliminary evidence that the C19-YRS outcome measure of PCS patients has satisfactory psychometric properties.

ACKNOWLEDGMENTS
NHS England clinical guidance suggests the use of the C19-YRS at first assessment, at six weeks, and at six months in PCS. 22