The number of patients “worse than death” while waiting for a hip or knee arthroplasty has nearly doubled during the COVID-19 pandemic

Introduction

The COVID-19 pandemic has had a significant impact on elective arthroplasty in the UK, with the cessation of nonurgent cases in mid to late March 2020.^1-3 When the incidence of new COVID-19 cases decreases across the UK the recommencement of planned orthopaedic elective services will need to be prioritized alongside other cancer and surgical services as resources may not be available for a full return to normal practice. When normal capacity is resumed there will be increased numbers of patients waiting much longer for hip and knee arthroplasty. There may also be some patients on the waiting list who may prefer to defer surgery due to the mortality risk associated with COVID-19 in the postoperative period, despite this being relatively low even at the height of first wave of the pandemic.⁴

Scott et al⁵ have previously shown the quality of life of patients waiting for hip and knee arthroplasty to be “worse than death” (WTD) for 19% and 12%, respectively, when defined as an EuroQoL (EQ) index of less than zero. However in their study, patients were prioritized according to their symptoms and surgery for those with the worst quality of life was expedited. During the COVID-19 pandemic and cessation of elective services, prioritizing those with the worst quality of life may have not been possible. Recommencement of services may lead to prioritization of patients who have the lowest surgical risk, which may further delay surgery in those with worst quality of life due to their associated comorbidity.¹ Currently there are no available data in relation to the quality of life of patients on NHS waiting lists for primary hip and knee arthroplasty. It is unknown whether patients would want to undergo surgery due to the associated risk of COVID-19.⁴

The primary aim of this study was to assess the quality of life of patients on the waiting lists for a hip or knee arthroplasty during the COVID-19 pandemic. The secondary aims were to assess: whether length of time on the waiting list influenced quality of life; patient opinions and options for future surgery; if patients wanted to proceed with surgery or delay intervention in view of the risks associated with COVID-19; the reason(s) for delaying surgery if this is their decision; and whether these metrics differed between patients listed for hip or knee arthroplasty.

Methods

A multicentre cross-sectional study was conducted across ten orthopaedic departments (Aberdeen, Bristol, Edinburgh, Fife, Glasgow, Leicester, Newcastle, Oswestry, University College London, and Wrightington) in the UK of patients on the NHS waiting lists for either a total hip (THA) or total (TKA) or partial knee arthroplasty (PKA) during the months of August and September 2020. This survey was registered and approved at each of the participating sites as an audit project or a quality improvement project. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for reporting observational studies were followed.⁶

Patients meeting the inclusion and exclusion criteria were selected from waiting lists at the ten study centres. When the survey began, six centres were participating with the aim to obtain data on 100 patients (50 THA and 50 KA) to meet the power calculation. However, four additional centres requested permission to participate during the study period and therefore 843 patients were included across ten sites. Patient demographic data (age and sex), proposed procedure (THA, TKA, or PKA), and date when listed were recorded. Patients were contacted by telephone and asked to complete an interviewer-administrated verbal EQ-5D questionnaire⁷ and in addition were asked questions relating to their quality of life, opinions about undergoing surgery in view of the COVID-19 pandemic and whether they wanted to defer their surgery and the reasons for this. The study protocol, hard copy patient questionnaire (should the researcher have required this while undertaking the telephone interview), and the data collection tool were sent to each site prior to the commencement of the study (Supplementary Material). The inclusion criteria were patients placed on the non-urgent elective orthopaedic waiting list for a primary THA or P/TKA prior to and including March 2020 who were able to complete a telephone interview. Patients listed for revision surgery or urgent arthroplasty were excluded. The chosen interviewer at each site was left to the centre’s discretion and availability of staff.

The EQ general health questionnaire evaluates five domains: mobility, self-care, usual activities, pain/discomfort, and anxiety/depression.⁷ The three-level (3L) version of the EQ-5D questionnaire was used, with responses to the five domains recorded at three levels of severity (no/slight problems; moderate/severe; or unable/extreme problems). Permission was obtained from the EuroQol Research Foundation to use the UK interviewer-administered version of the EQ-5D-3L version. This index is on a scale of -0.594 to 1, where 1 represents perfect health and 0 represents death.⁸ Patients scoring less than zero for the EQ-5D score were defined to be in a state WTD.⁵ The minimal clinically important difference (MCID) in the EQ-5D score after THA is 0.08, therefore a difference in the score of 0.08 or more was defined as clinically important.⁹ The EQ visual analogue scale (VAS) was also completed, again using the UK interviewer-administered version, that assesses how good or bad the patient’s health is on that day and ranges from 100 (best health) to 0 (worst health).

Data from the previously published study by Scott et al⁵ were used for the unmatched control groups and were thought to represent baseline data over a period (from January 2014 to September 2017) that was not affected by the COVID-19 pandemic. Sex, age, EQ-5D, and EQ-VAS data were available for 2,073 patients waiting for a THA and 2,168 patients waiting for a TKA.

Results

During the study period, 843 patients reported their EQ-5D and EQ-VAS scores and completed the COVID-19 questionnaire. There were 344 male (40.8%) and 499 female patients (59.2%) with a mean age of 69.1 years (26 to 98). There were 394 waiting for a THA and 449 waiting for a TKA (n = 418) or a PKA (n = 31). For analytical purposes, TKA and PKA were grouped together as a knee arthroplasty (KA) group.

Quality of life of patients currently on the waiting list

There were no significant differences in age or sex between the 2020 group and the control groups (Table I). The 2020 groups had significantly (p < 0.001, independent-samples t-test) worse EQ-5D and EQ-VAS scores compared with the control groups, for both patients awaiting THA and KA (Table I and Figure 1). Over one-third (35.0%; n = 138/394) of patients waiting for a THA and nearly a quarter (22.3%; n = 100/449) of patients waiting for a KA were in a state WTD, which was approaching twice that observed in the control groups for both THA (p < 0.001, chi-squared test) and KA (p < 0.001, chi-squared test) (Table I). There was no significant difference (p = 0.505, Fisher’s exact test) between TKA and PKA for the number of the patients WTD (n = 95 (22.7%) vs n = 5 (16.1%), respectively). The worse health-related quality of life in the 2020 groups relative to the control groups is illustrated by a shift to the left in the density plots of the EQ-5D scores (Figures 2 and 3).

Fig. 1

Bar chart of EuroQol five-dimension (EQ-5D) scores for the control (pre-COVID-19) and COVID-19 (2020) groups for patients waiting for a total hip (THA) or knee (KA) arthroplasty. The error bars represent 95% confidence intervals. *p < 0.001, independent-samples t-test.

Fig. 2

Density plot for the distribution of EuroQol five-dimension (EQ-5D) scores for patients awaiting a total hip arthroplasty before the pandemic and for the 2020 cohort.

Fig. 3

Density plot for the distribution of EuroQol five-dimension (EQ-5D) scores for patients awaiting a knee arthroplasty before the pandemic and for the 2020 cohort.

Fig. 4

Scatter plot and line of best fit for correlation between the time patients were listed for their total hip (THA) (grey, r = 0.12) or knee (KA) (black, r = 0.10) arthroplasty and mean EuroQol five-dimension (EQ-5D) score in August/September 2020.

Table I.

Demographic data, EuroQol five-dimension, EuroQol visual analogue scale, and worse than death status for total hip and knee arthroplasty according group.

Variable	Control	2020	Difference/OR (95% CI)	p-value
THA	(n = 2,073)	(n = 394)
Mean age, yrs (SD)	67.4 (11.6)	68.0 (12.3)	0.7 (-0.6 to 1.9)	0.304*
Sex, n (%)
Female	1,253 (60.4)	240 (60.9)	Reference
Male	820 (39.6)	154 (39.1)	0.98 (0.79 to 1.22)	0.861†
Mean EQ-5D (SD)	0.360 (0.325)	0.241 (0.351)	0.119 (0.083 to 0.154)	< 0.001*
Mean EQ-VAS (SD)	67.4 (22.2)	54.0 (23.3)	13.3 (10.8 to 15.9)	< 0.001*
WTD, n (%)
No	1,682 (81.1)	256 (65.0)	Reference
Yes	391 (18.9)	138 (35.0)	2.30 (1.83 to 2.93)	< 0.001†
KA	(n = 2,168)	(n = 449)
Mean age, yrs (SD)	69.4 (9.6)	70.0 (9.4)	0.7 (-0.3 to 1.7)	0.159*
Sex, n (%)
Female	1243 (57.3)	259 (57.7)	Reference
Male	925 (42.7)	190 (42.3)	0.99 (0.80 to 1.21)	0.891†
Mean EQ-5D (SD)	0.408 (0.311)	0.335 (0.327)	0.074 (0.042 to 0.105)	< 0.001*
Mean EQ-VAS (SD)	69.8 (20.6)	58.3 (21.9)	11.4 (9.2 to 13.7)	< 0.001*
WTD, n (%)
No	1,906 (87.9)	349 (77.7)	Reference
Yes	262 (12.1)	100 (22.3)	2.08 (1.61 to 2.70)	< 0.001†

1. *

   Independent-samples t-test.

1. †

   Chi-squared test.

1. CI, confidence interval; EQ-5D, EuroQol five-dimension score; EQ-VAS, EuroQol visual analogue scale; KA, knee arthroplasty; OR, odds ratio; THA, total hip arthroplasty; WTD, worse than death.

Discussion

This study has shown that patients waiting for THA and KA during the COVID-19 pandemic had a significantly worse quality of life than expected relative to previous years. Over one-third of patients waiting for a THA and nearly one-quarter of patients waiting for a KA were in a health state WTD, which was approaching double that observed prior to the pandemic. There was a direct correlation between increasing length of time of the waiting list and a worsening quality of life: for each additional six months of waiting there was a clinically significant deterioration in health-related quality of life. This was supported subjectively by the patient's opinion, with over 80% stating that their quality of life had deteriorated while waiting for their arthroplasty. The majority of patients (86.1%) were theoretically willing to undergo their surgery despite the potential risks associated with COVID-19; however, half wished to have a face-to-face consultation should they go forward with surgery. Half the patients were happy to have their arthroplasty under a different surgeon or hospital. Those wishing to defer surgery had concerns for themselves and or their family and were more likely to be waiting for a KA.

The EQ-5D score has been proven to be a reliable and responsive measure of health-related quality of life and is validated for use in a number of populations and conditions including degenerative joint disease.^10,11 A total of 243 health states (indices) are possible and differ by country according to population preferences. This UK-based study uses the UK value set in which 84/243 (35%) of the possible health states are negative and can be defined as WTD.¹² This WTD definition is researcher-generated and reflects the construction of the EQ-5D score.¹⁰ These negative or WTD states have been defined as such by 3,395 individuals from the UK grading each state using the time trade-off method: the less time in the health state that could be endured, the lower the score.¹² These negative (WTD) scores are therefore a hypothetical value judgement made by the UK general public based on their beliefs of health-related quality of life. The ability to score negative health states enables a broader description of severe disability and severely poor health, and lessens the floor effect of this score. However, despite having a WTD score 28 patients did not want to go forward with surgery due to fears of contracting COVID-19 in hospital, which may not fully support their WTD status or may reflect patient choice to delay surgery until they perceive it is safer to proceed. The control cohort of patients used in the current study demonstrated that using this definition, in normal circumstances, 19% of patients waiting for THA and 12% of patients awaiting TKA were in a health state WTD.⁵ Under normal circumstances, these patients who are the worst affected by degenerative joint disease would be prioritized and thus tend to wait less time for their surgery.⁵ It has therefore not been previously possible to investigate the effect of delayed and prolonged waiting times on health-related quality of life.

At the start of the COVID-19 pandemic all non-urgent operating within the NHS was suspended. Trauma and orthopaedics were reduced to just 3.3% of normal operating volumes in England, the lowest percentage of any surgical speciality.³ Since restarting after the first wave, rates of orthopaedic admissions for surgery peaked at 63% of normal in England³ and in Scotland arthroplasty surgery was being performed at a rate of 40% of normal.² This reduction of 60% in operating capacity far exceeds the 14% of patients who would have potentially chosen to defer their surgery currently because of the pandemic. As the pandemic has continued, the number of patients who wish to defer surgery is reducing.^4,13 This may reflect a reduction in the fear of COVID-19 as vaccines are developed, but similarly could reflect a deterioration in health-related quality of life altering the balance of perceived risk.⁴

COVID-19-free pathways have enabled some elective orthopaedic operating during the pandemic and have been shown to be effective and safe in terms of both viral transmission and morbidity and mortality.^13,14 Developing COVID-19 during the perioperative period is associated with excess mortality;^15,16 however, the postoperative mortality risk due to COVID-19 following hip and knee arthroplasty at the peak of the first wave of the pandemic has been estimated at 1/1,000.¹⁵ As a speciality, the current rate of surgery in orthopaedics is lower than in all other surgical specialties except oral surgery.¹⁷ In England the number of patients waiting for surgery is higher for orthopaedics than for any other surgical speciality.³ Using data from the Scottish Arthroplasty Project, Yapp et al² have reported that had arthroplasty operating been restarted at a rate of 120% of pre-COVID activity in October 2020 it would take 24 to 27 months to clear the backlog. This is also supported by waiting list data from England that estimated with 30% increased activity it would take 20 months if there was no hidden burden of unreferred patients, and 48 months if there was a hidden burden, to return to pre-COVID-19 waiting list numbers.¹ However, it may not be possible to work at 120% or 130% within the NHS for the foreseeable future. Presently additional capacity from the private sector may not be available in the future due to pressures on their own waiting lists. Non-urgent elective operating has again been postponed due to the second wave of the pandemic and therefore this waiting backlog is currently increasing.

The current study has demonstrated not only that a third of THA patients and nearly one-quarter of KA patients are now living in a health state WTD, but also that longer time on the waiting list was associated with a significantly worse health-related quality of life. As of December 2020, approximately 40,000 patients in England and Scotland had been waiting over one year for orthopaedic surgery.^3,18 Based on previous levels of arthroplasty, approximately one-third of these patients are likely to be waiting for hip or knee arthroplasty.¹⁸ This translates to more than 13,000 patients waiting over one year after listing for hip or knee arthroplasty. While patients wait longer their pain is managed in primary care where medical practitioners may come under increasing pressure to prescribe opiate medication, against current recommendations for osteoarthritis management.¹⁹ The current study has found that clinically significant deteriorations exceeding the EQ-5D MCID occur with each additional six-month wait for hip or knee arthroplasty. Similar to Morris et al,²⁰ more than 80% of patients subjectively reported their health-related quality of life had deteriorated while on the waiting list during the pandemic. EQ-5D scores in patients with degenerative joint disease of the hip and knee are known to be independently associated with joint-specific patient-reported outcome measures (PROMs) and thus pain and function from the involved joint.⁵ It has also previously been demonstrated that patients with worse preoperative EQ-5D scores ultimately achieve worse Oxford Hip and Knee Scores and worse rates of patient satisfaction following hip and knee arthroplasty⁵ and are less likely to return to work.²¹ Delaying these patients on indefinite waiting lists and deprioritizing the recommencement of “elective” orthopaedic operating is not benign: it will have real and lasting effects on these patients who may achieve poorer outcomes as a result.

This study has limitations. Comorbidities and joint-specific PROMs were not included. This was for ease of applying the questionnaire over the telephone. The relationship between comorbidities, joint-specific PROMs, and EQ-5D indices have been investigated and delineated previously in a large cohort.⁵ Resource allocation within a healthcare setting like the NHS is based on quality adjusted life years that are derived from health utility scores such as the EQ-5D score, not on joint-specific function. Therefore, the EQ scores were collected in isolation. Patients were selected at random from the waiting lists to avoid selection bias but were not consecutive and non-responders may have influenced the interpretation of results. The multicentre nature of the study however enhances its generalizability and external validity. The comparative control cohort from 2014 to 2017 was from a single centre, including over 4,000 arthroplasty patients with nearly identical preoperative EQ-5D scores to that observed in the National Joint Registry, and is therefore likely to be representative of patients awaiting hip and knee arthroplasty across the UK.^22,23 Furthermore, the single centre providing the control group data was not significantly different from the other centres providing data for the 2020 group (Table II). However, those in the 2020 group were interviewed during a global pandemic and while during August and September 2020 in the UK there were only limited social restrictions in place, this was a different landscape to that of the control data collection period of 2014 to 2017. Therefore, it is not possible to be certain whether a societal state of pessimism due to the COVID-19 pandemic influenced the patients' perceptions when they responded to the questionnaire or if a real physical deterioration had occurred. However, without sampling people not on the waiting list to assess the influence of COVID-19 restrictions on the population, this persists as a limitation.

In conclusion, over a third of patients waiting for a THA and nearly one-quarter of patients waiting for knee arthroplasty were in a health state WTD, almost double that observed previously. Subjectively over 80% of patients reported a decline in their health-related quality of life while on the waiting list. Objectively health-related quality of life scores correlated and declined significantly with time on the waiting list and therefore those in a state WTD will likely grow as the wait time increases. In a societal context this is highly relevant, especially when considering the health service rationing and prioritization that will be required not only during the recommencement of routine operating but over future years as the NHS recovers from the impact of COVID-19.