Socioeconomic Area Deprivation is Related to Poorer Patient-Reported Outcomes Following Autologous Chondrocyte Implantation (ACI)

Abstract

Objective

To determine the association between socioeconomic deprivation and short-term patient-reported clinical outcomes following autologous chondrocyte implantation (ACI).

Design

All patients receiving knee ACI between 1996 and 2020 in our center were identified. Socioeconomic deprivation of their residential area was quantified using the Index of Multiple Deprivation (IMD). Patient-reported 1-year Lysholm and Intermittent and Constant Osteoarthritis Pain (ICOAP) scores were used as outcome measures in the analyses. After transformation to ensure normal distributions (where required), linear multivariable regression was used to analyze the relationship between IMD and 1-year Lysholm score, adjusting for demographic characteristics (age, sex, body mass index [BMI], and smoking) and baseline Lysholm.

Results

Three hundred and ninety-one patients with a mean age of 50 years (range = 16-84; 266 male) were identified. Median BMI was 27 (17-47), with 138 patients overweight and 105 obese. Seventy-seven patients lived in upper and 41 in lower quintile deprivation areas. The mean baseline Lysholm score was 49.8 ± 17.3 SD, improving to 66.5 ± 21.3 SD at 1 year. Mean 1-year Lysholm scores were significantly lower with increasing area deprivation scores, adjusted for demographic factors. Specifically, areas with high unemployment levels, being female, or having a lower baseline Lysholm were associated with poorer outcomes, but age, BMI, smoking, or higher income deprivation were not.

Conclusion

This study demonstrates poorer functional outcomes following ACI in patients from more deprived areas, indicating future studies should consider neighborhood deprivation as a confounding factor. Furthermore, targeting patients from areas with higher deprivation with additional interventions/community support may improve their outcomes.

Keywords

autologous chondrocyte implantation socioeconomic deprivation BMI rehabilitation patient-reported outcome

Introduction

Population health needs are the main driving force in any health service provision, including the UK’s National Health Service (NHS).¹ Socioeconomic status (SES), a composite measure related to social structure based on an individual’s education, income, and occupation concerning others.² SES is widely recognized as a fundamental determinant of health care disparities and is a critical issue affecting all fields of medicine. A patient’s SES can influence their requirements for access to health care services. For example, patients from disadvantaged backgrounds and deprived areas visit primary care providers more frequently than those from more affluent backgrounds, consistent with their poorer general health and a greater need for health care provision. Individuals from lower-income areas are also less likely to receive outpatient care³ and experience less favorable recovery from surgery.^4,5

The British government is working on many projects targeting patients from deprived areas and channeling more resources to tackle such health inequalities and improve health care outcomes for individuals from deprived areas. For example, Core20PLUS5 is the NHS England approach to tackle health inequalities among the most deprived 20% of the population (as defined by the Index of Multiple Deprivation [IMD]) in 5 defined clinical focus areas. For adults, this includes maternity, severe mental illness, chronic respiratory disease, early cancer diagnosis, and hypertension case-finding, and for children and young people, this includes asthma, diabetes, epilepsy, oral health, and mental health.⁶

Focal articular cartilage defects of the knee can cause significant pain and disability and ultimately lead to degenerative arthritis if left untreated. Autologous chondrocyte implantation (ACI), first published as a restorative treatment option 3 decades ago, is a 2-step surgical procedure that involves arthroscopically removing autologous chondrocytes from a low weightbearing area of the patient’s knee for culture expansion in a GMP-licensed laboratory, followed by the implantation of cells into the symptomatic defect in the same knee approximately 2 to 4 weeks later.⁷

Compared with alternative techniques, ACI provides good mid- and long-term clinical outcomes^8
-11 perhaps due to maturation¹² with time in both the chemical and structural properties of the repair tissue,^13
-15 as well as the repair matrix becoming biomechanically stiffer.¹⁶ ACI has been shown to reduce pain in active young populations and improve knee function by returning athletes to their preinjury activity level at rates of between 73% and 84%.¹⁷ Clinical results post-ACI are affected by several variables, such as patient characteristics, lesion characteristics, surgical repair effectiveness, and postoperative recovery.¹²

With regard to orthopedics, the quality of care provided, whether by surgical intervention or nonoperative treatment, is closely intertwined with social factors related to the SES of a patient.¹⁸ Within orthopedic surgery, SES is a proven predictor of morbidity and mortality; hence, modifiable variables contributing to low SES must be assessed and maximized to lessen health care inequalities.¹⁹ With the emergence of value-based health care and solutions seeking to improve outcomes and reduce cost, progress is being made toward reducing negative surgical outcome predictors associated with SES.¹⁸

To the best of our knowledge, no study has previously investigated the impact of socioeconomic factors on the functional outcomes of patients undergoing ACI procedures. Predictors of ACI outcomes have focused primarily on surgical technique, implant/defect characteristics, and individual patient clinical factors.^20,21 The relative importance of these factors and their potential interactions with the patient’s SES are still unknown. Therefore, in this study, we determined the social deprivation levels of patients undergoing ACI knee surgery and assessed their association with clinical outcomes.

Methods and Materials

Data Collection and Handling

A retrospective analysis was performed based on all patients who had ACI surgery in our center between 1996 and 2020 with a minimum of 12 months’ follow-up. Patients from various socioeconomic backgrounds across the United Kingdom receive ACI in our center, predominantly from the Northwest, West and East Midlands in England, and Mid/North Wales. This study was approved as a service evaluation by our in-house Clinical Audit Department.

Baseline demographics including postcode, age, gender, smoking habits, height (m), and weight (kg) were collected, and body mass indices (BMIs) were calculated. Patients with a BMI of 20 to 24.9 kg/m² were categorized as having a normal weight, those with a BMI of 25 to 30 kg/m² as being overweight, and those with a BMI >30 kg/m² as obese. Two patient-reported outcome measures (PROMs) were collected—the Intermittent and Constant Osteoarthritis Pain score (ICOAP)²² and the Lysholm score at baseline and 12 months postoperatively.^23,24

The Lysholm score is a 100-point scoring system (100 represents a perfectly functioning knee) for examining a patient’s knee-specific symptoms and function, including mechanical locking, instability, pain, swelling, stair climbing, and squatting. In this study, we used a modified version that fitted a previously published Rasch model analysis, by omitting the swelling item and giving the remaining 7 items similar weights.²⁴ The ICOAP is an 11-item questionnaire divided into 2 domains: a 5-item scale for constant pain and a 6-item scale for intermittent pain (so-called “pain that comes and goes”). The pain score is rated by pain intensity, frequency, impact on mood, sleep, and quality of life.

We assessed SES using the IMD,²⁵ which provides a well-established area-based relative measure of deprivation for small areas or “Lower-layer Super Output Areas” (LSOA) across England, based on 7 domains of deprivation: Income Deprivation, Employment Deprivation, Education, Skills and Training Deprivation, Health Deprivation, Disability, Crime, Barriers to Housing and Services, and Living Environment Deprivation. The domains were combined using their default weights to produce the overall IMD score (range 0 to 100, 0 = least, 100 = most deprived). Patients’ postcodes were used to derive the IMD scores for the area where the patient lived at the time of surgery; patients were then placed into 1 of 5 SES groups based on quintiles of the national distribution of IMD scores. The IMD quintile group 1 represents patients living in the 20% highest ranked (most deprived) areas, and IMD quintile group 5 represents those in the 20% lowest ranked (least deprived) areas. As some of the indices have changed over time, the index closest to the operation date was used.

Statistical analysis

Baseline, 12-month, and latest Lysholm scores were compared using paired t-tests. The 5 IMD quintile subgroups were compared in terms of smoking behavior and sex using Fisher’s exact test, in terms of median BMI using the Kruskal-Wallis test, and in terms of remaining mean demographic characteristics and clinical outcomes using 1-way analysis of variance (ANOVA).

Multivariable linear regression was used to explore associations between socioeconomic deprivation and the postoperative PROMs (Lysholm and ICOAP). In these analyses, 2 separate models were used, with either the IMD index total score or 2 IMD individual domain scores, namely employment and income, as independent variables. Four potential confounders (sex, age and 2 lifestyle factors, BMI, and smoking status) were accounted for by including them as covariates. Incomplete data were analyzed using the full information maximum likelihood.²⁶ Distributions were checked using quantile-quantile (QQ) plots, and a transformation from the Box-Cox power family was applied if needed. If a suitable transformation could not be found, a robust 2-stage approach for the full information maximum likelihood (FIML) was undertaken.²⁷ Patients could answer the question of smoking habits with “yes,” “no,” or “ex,” where no specific cutoff point in time was provided for “ex.” Because the effect of being an ex-smoker depends strongly on the time-frame involved, we decided to regard “ex” as a missing variable.²⁸ All analyses were performed using R vs 4.0.5 software (R Foundation for Statistical Computing, Vienna, Austria) using the packages “lavaan” and “semTools.” A 2-tailed P-value below 0.05 was assumed to denote significance.

Results

A total of 391 patients (266 male, 125 female) were identified. Age and Lysholm scores were approximately normally distributed, but BMI and the various deprivation indices required log transformation. The mean age of the patients was 50 years (range = 16-84), with a median BMI of 28 (range = 17-47) and a median IMD of 14 (range = 1.9-73.4; Table 1 ). There was a significant variation in mean age across the deprivation quintiles, with patients living in more deprived areas having a lower mean age (P = 0.031; Table 2 ). We found no evidence that the sex distribution or median BMI varied between the IMD quintiles ( Table 2 ). Of those answering the question regarding smoking status, 16% stated they were smokers at the time of operation and 4% stated they were ex-smokers, with no evidence that this percentage varied among the deprivation quintiles ( Tables 1 and 2 ).

Table 1.

Demographic and Clinical Characteristics of Entire Cohort.

Variable	Mean (SD), Median [IQR], or n (%)	Range	Total Number	Number with Missing Info
Age	49.6 (11.6)	16-84	391	0
Male/female	266/125 (68/32%)		391	0
BMI	27 [25 to 30]	17-47	321	70
Smoker (N/Y/Ex)	185/37/8 (80/16/4%)		230	161
IMD score	14.4 [8.1 to 23.9]	1.9-73.4	293	98
Preop Lysholm	49.8 (17.3)	12.5-100	346	45
1-year Lysholm	66.5 (21.3)	16.7-100	259	132
Latest Lysholm	63.7 (23.3)	8.3-100	331	60
Total pain	20.5 [6.8 to 45.5]	0-100	253	138

Table 2.

Demographic and Clinical Characteristics, Divided by IMD Quintile.

	IMD Quintile
Variable	1	2	3	4	5	P
Total (n)	41	46	67	62	77
Sex (n, %)						0.61^a
Female (n = 125)	13 (32)	17 (37)	20 (30)	24 (39)	21 (27)
Male (n = 266)	28 (68)	29 (63)	47 (70)	38 (61)	56 (73)
Age (mean (SD))	46.0 (10.7)	45.4 (11.1)	50.6 (10.6)	51.1 (11.2)	49.8 (13.6)	0.031 ^b
BMI (median [IQR])	30 [25, 32]	29 [24, 31]	27 [24, 29]	28 [24, 30]	27 [25, 30]	0.16^c
Smoker (n, %)
No	22 (92)	19 (70)	34 (85)	33 (85)	45 (83)	0.19^a
Yes	2 (8)	8 (30)	3 (7.5)	4 (10)	8 (15)
Ex	0	0	3 (7.5)	2 (5)	1 (2)
Preop Lysholm (mean (SD))	50.6 (20.6)	46.6 (15.1)	50.8 (17.7)	49.1 (16.9)	52.4 (17.6)	0.53^b
Total pain (median [IQR])	36.4 [22, 57]	20.5 [0, 43]	11.4 [5, 27]	15.9 [7, 47]	28.2 [7, 38]	0.092^c
Twelve-month Lysholm (mean (SD))	56.2 (22.0)	71.3 (20.3)	67.4 (24.2)	68.6 (17.8)	72.0 (19.5)	0.024 ^b

Variables with significant differences across quintiles indicated using bold typeface.

Fisher’s exact test.

One-way ANOVA.

Kruskal-Wallis.

The mean preoperative Lysholm score was 50 (range = 12.5-100) and improved significantly to a mean score of 66.5 (range = 16.7-100) 12 months postoperatively (P < 0.001; Table 1 ). No evidence was found that the mean baseline Lysholm score varied across the deprivation quintiles, but the mean 12-month score varied significantly, with quintile 1 (most deprived) showing the lowest mean (P = 0.024; Table 2 ). By the latest follow-up (at a median of 10 years), the mean postoperative Lysholm score had marginally reduced to 64 (range = 8.3-100), but the change from the 12-month results was not significant (P = 0.13). No evidence was found that the median total pain score varied across the IMD quintiles.

Multivariable Analyses Adjusted for Potential Confounders

No evidence was found that the preoperative outcome score (Lysholm) was associated with the local area IMD score (P = 0.53) or its domains of employment (P = 0.22) and income (P = 0.095) when adjusted for potential confounders (sex, age, BMI, smoking; Table 3 ). As regards confounders, male patients had a statistically higher mean preoperative Lysholm score than their female counterparts (P < 0.001, all models), and patients who were smokers (P < 0.001, all models) or had higher BMI (P = 0.048, IMD model) had lower mean baseline Lysholm scores.

Table 3.

Multivariable Regression Analyses for Baseline Lysholm Score.

	Model 1 (IMD Score)		Model 2 (Employment and Income)
Variable	Coefficient (95% CI)	P	Coefficient (95% CI)	P
Male	10.95 (7.27 to 14.63)	<0.001	10.66 (6.99 to 14.34)	<0.001
Age	0.023 (-0.13 to 0.17)	0.76	0.0070 (-0.14 to 0.16)	0.92
Smoker	-25.83 (-37.82 to -13.84)	<0.001	-25.14 (-37.21 to -13.080)	<0.001
log(BMI)	-6.22 (-12.38 to -0.062)	0.048	-5.82 (-12.018 to 0.37)	0.065
log(IMD)	-0.80 (-3.32 to 1.72)	0.53	-	-
log(employment)	-	-	4.34 (-2.52 to 11.19)	0.22
Log(income)	-	-	-4.46 (-9.70 to 0.77)	0.095

Regression coefficients, their 95% CIs, and P-values based on a structural equation model using the full information maximum likelihood (FIML). Covariates with significant associations across both models are indicated using bold typeface.

Twelve-month postoperative Lysholm scores did show a statistically significant association with IMD scores (P = 0.016), with patients from a more deprived area (higher IMD) having lower outcome scores ( Table 4 ). A similar relationship was found between area unemployment level and 12-month Lysholm score, with patients from high unemployment areas having worse outcomes (P = 0.008). Regarding confounders, mean 12-month Lysholm scores were higher among male patients (P = 0.036, IMD model and P = 0.017, employment and income model) and among those who had higher baseline Lysholm scores (P < 0.001, all models). There was no statistical evidence for an association between 12-month Lysholm score and income (P = 0.17), age (P = 0.46, IMD model and P = 0.42, employment and income model), BMI (P = 0.61, IMD model and P = 0.60, employment and income model), or smoking (P = 0.30, IMD model and P = 0.30, employment and income model).

Table 4.

Multivariable Regression Analyses for 12-Month Lysholm Score.

	Model 1 (IMD Score)		Model 2 (Employment and Income)
Variable	Coefficient (95% CI)	P	Coefficient (95% CI)	P
Male	5.67 (0.37 to 10.97)	0.036	6.42 (1.16 to 11.69)	0.017
Age	-0.079 (-0.29 to 0.13)	0.46	-0.086 (-0.30 to 0.12)	0.42
log(BMI)	-4.40 (-21.27 to 12.48)	0.61	-4.41 (-21.093 to 12.27)	0.60
Smoking	-4.44 (-12.79 to 3.91)	0.30	-4.26 (-12.52 to 3.99)	0.31
Preop Lysholm	0.40 (0.25 to 0.55)	<0.001	0.41 (0.26 to 0.56)	<0.001
log(IMD)	-4.49 (-8.16 to -0.82)	0.016	-	-
log(employment)	-	-	-12.16 (-21.16 to -3.16)	0.008
Log(income)	-	-	4.94 (-2.088 to 11.96)	0.17

No evidence was found for an association between the total pain score and either the deprivation score (IMD) or 2 of its constituents (income and employment) when adjusted for potential confounders (sex, age, smoking, and BMI; Table 5 ). However, female patients (P = 0.021, IMD model and P = 0.019 employment and income model) and smokers (P = 0.02, IMD model and P = 0.022, employment and income model) reported significantly higher mean total pain levels.

Table 5.

Multivariable Regression Analyses for 12-Month Total Pain Score.

	Model 1 (IMD Score)		Model 2 (Employment and Income)
Variable	Coefficient (95% CI)	P	Coefficient (95% CI)	P
Male	-7.43 (-13.74 to -1.11)	0.021	-7.63 (-14.00 to -1.26)	0.019
Age	-0.063 (-0.33 to 0.20)	0.64	-0.032 (-0.30 to 0.23)	0.82
log(BMI)	7.23 (-16.97 to 31.44)	0.56	6.06 (-18.58 to 30.70)	0.63
Smoker	14.29 (2.22 to 26.36)	0.020	13.97 (2.03 to 25.90)	0.022
log(IMD)	3.11 (-1.98 to 8.21)	0.23
log(employment)			7.89 (-3.86 to 19.64)	0.19
log(income)			-1.31 (-11.40 to 8.68)	0.80

Regression coefficients, their 95% CIs, and P-values based on a structural equation model using the two-stage robust full information maximum likelihood (FIML). Covariates with significant associations across both models are indicated using bold typeface.

Discussion

This study describes a series of analyses of patient outcome data from a unique cohort of individuals who had received ACI treatment surgery in one of the major centers for this operation in the United Kingdom. The focus of the study was to investigate whether SES was associated with baseline or postoperative patient-reported outcome of ACI. Most notably, patients living in areas with higher levels of socioeconomic deprivation reported poorer functional outcome (Lysholm scores) 12 months post-ACI despite a lack of an association with outcome at baseline. Moreover, the observed association between higher deprivation and lower functional 12-month outcome was observed regardless of adjusting for potential confounding factors, namely preoperative functional score, age, gender, BMI, and smoking status.

Additional modeling analyses suggested that the association between SES and ACI outcome may be related specifically to patient income/employment, as lower income and employment levels were also associated with a poorer 12-month Lysholm. However, no associations between either of the investigated SES metrics and baseline Lysholm scores were observed. To our knowledge, no other studies have reported on SES associations among ACI patients, but studies of SES among anterior cruciate ligament (ACL) reconstruction patients have similarly reported that patient satisfaction was strongly influenced by employment status, with jobless patients reporting considerably lower satisfaction levels.^29,30 Reasons for the specific associations in our study are as yet unclear and will require further investigation, although 1 explanation for the association of SES with postoperative but not with baseline outcome could be related to the importance of the rehabilitation program for the outcome.¹² Engagement with or adherence to the lengthy rehabilitation programs recommended for this type of surgery could be reduced in areas that are deprived or have high unemployment due to limited access to rehabilitative health care facilities^29,30 influenced by different factors. What can exacerbate the problem is the lack of community-based rehabilitation resources and programs, leaving the patients without sufficient support to help in their recovery.

Support for this suggestion comes from a large population-based survey among stroke patients in South Korea in 2023, which found direct evidence that accessibility to physiotherapy was affected by patient SES, in addition to confounders such as the type of residential area (urban vs rural), patient age, and gender.³¹ Access to rehabilitation facilities, especially transportation, can significantly hinder patient compliance. Patients, especially in deprived areas, may find it challenging to attend physiotherapy sessions consistently, often due to reasons related to the inability to transport or affordability of the associated costs.^32,33 Another factor is employment constraints and the lack of flexibility in attending rehabilitation sessions due to working patterns and multiple jobs, which can lead to missed sessions.³³ What can exacerbate the problem is the lack of community-based rehabilitation resources and programs, leaving the patients without sufficient support to help in their recovery.³³ Language barriers, inadequate health literacy, and limited access to information about rehabilitation options can negatively affect motivation and engagement in the recovery process.³⁴

Related specifically to rehabilitation following sports injuries, a recent scoping review on the role of psychological, social, and contextual factors (the latter including SES) across recovery stages identified 77 studies, mostly (84%) focusing on ACL injuries, all (100%) addressing psychological factors, but only a small number (21%) addressing contextual factors, none of them socioeconomic factors.³⁵ The review therefore calls for more studies considering the broader systemic aspects of contextual conditions in which athletes are injured and recover.³⁶ Results from such research can then be used to change the practice of physiotherapy and rehabilitation.³² The results from the current study can only make suggestions around the link between SES and rehabilitation following ACI, but a randomized controlled trial (RCT) of cell therapy at our institution does include a study of physiotherapy adherence and will be able to expand on this proposed link.³³

When analyzing the influence of SES, we controlled for commonly used potential confounders (age, sex, BMI, and smoking status) that could affect ACI outcome but are also linked to SES.³⁷ This group of patients did not show clear differences in values for these confounders across IMD quintiles, other than patients from more deprived areas having a lower mean age. Adjusting for these confounders therefore did not affect the results of our analysis.

In terms of the relationship between these 4 confounders and ACI outcomes, our findings were mostly in line with existing studies. Our study found no connection between patient age and functional outcome, like previous studies among which 3 RCTs that included numerous patients above the age of 50 years.^38
-40 The current study did find higher mean baseline and 12-month functional outcomes in males and increased pain postoperatively in females. The effect of patient sex on PROMs following ACI surgery varies across the literature. Some showed that being female was associated with poorer outcomes, which mirror the current study.^41,42 However, other studies, including a systematic review and meta-analysis, suggest that sex does not predict outcomes following ACI surgery.^38,43

The BMI and smoking status were also considered as confounders. Our study found that higher BMI was associated with lower baseline functional scores but not with 12-month outcomes. However, the latter finding is simply the effect of controlling for baseline Lysholm score when analyzing the 12-month score, i.e., we found no additional effect of smoking above that imparted via the baseline score. Relatively few studies have investigated the relationship between level of obesity and ACI outcomes. A study of predictive factors for long-term outcomes following Matrix-induced Autologous Chondrocyte Implantation (MACI) found no evidence that BMI affects outcome.¹² However, this study also controlled for baseline score (the most well-established predictor), which probably explains why BMI had no additional predictive value. On the contrary, a case series comparing normal, overweight, and obese patients found that obese patients had lower mean Modified Cincinnati knee scores before surgery, similar to our findings, but also showed that they had not any benefit from ACI or MACI 2 years postsurgery.⁴⁴ In addition, a linkage study with the Norwegian arthroplasty registry showed that a higher BMI increases the risk of total knee replacement surgery post-ACI, although this study only assessed BMI at the mean 20-year follow-up.⁴⁵ More research is needed to explore the potential longitudinal BMI associations with outcomes following ACI and other orthopedic surgeries.

Similar to our findings on BMI, we found that smoking had a negative effect on baseline ACI outcome but not on 12-month outcome. The impact of smoking on cartilage procedures remains controversial with conflicting outcomes from different studies over the years. Jaiswal et al. showed in their case-control study that smoking was associated with poorer baseline PROMs, similar to the current study, and increased pain post-ACI surgery.^46,47 However, a recent study of the German cartilage registry concluded that smoking did not affect the outcome within the first 2 years after third-generation ACI.⁴⁸ The same finding of no impact of smoking status on outcomes at 12 months post-ACI was replicated in the current study, although this is probably related to our controlling of 12-month outcomes for the baseline Lysholm score.

We have provided unique and novel socioeconomic data on the largest UK ACI cohort. However, a few limitations need to be considered when interpreting the results of the current study. The level of missing data in the study represents an obvious limitation but is often unavoidable in retrospective analyses of observational data, even if the data are collected prospectively. However, multivariable regression using full information matrix maximum likelihood (FIML) was used to handle missing data where required, a technique that can be used successfully even with large proportions of missing data.²⁶ The technique assumes that data are either missing completely at random (MCAR; missingness is independent of the observed data and the missing data itself) or missing at random (MAR; missingness partly dependent on the observed data but not the missing data itself). If, however, data were missing not at random (MNAR; missingness depends on the data itself), for instance, if patients with poorer outcome were less likely to report their outcome, the result may be biased.⁴⁹ If the 132/391 (34%) missing 12-month scores were mainly due to poorer outcome, the risk of bias would be large. However, a systematic review identified 46 factors associated with missing PROMS, of which only 3 (physical impairment, disease progression, and level of wellbeing) might give a negative bias.⁵⁰ Follow-up studies of nonresponding patients who had hip or knee arthroplasty or spine surgery suggest that PROMs do not differ significantly between responders and nonresponders, suggesting the other 43 factors are more important.^49,51 Nonresponse of hip arthroplasty patients was particularly associated with younger age, lower baseline PROMs, higher BMI, and social deprivation, all 4 included in our analysis^51,52 This suggests the MAR assumption in our analysis was correct and therefore the risk of bias small. Another limitation is that the study was a population-based study. Given that deprivation is thought to have the most significant impact on individuals,⁵³ quantifying socioeconomic deprivation using an area-based measure may understate the relationship between deprivation and PROMs and result in some degree of individual misclassification. Furthermore, our study used the IMD score, a nationally recognized public health tool for measuring social deprivation and SES. The IMD comprises 7 domains, mainly income, and employment, making it a reasonable tool to reflect SES in a small geographical area of 1,500 people each. We need to acknowledge the fact. However, this implies that IMD scores reflect ecological community deprivation status instead of reflecting individual SES variation. In a cohort of cancer patients in England and Wales, the concordance between the (area-based) indices of income, education, and occupation versus their individual-based counterparts was poor.⁵⁴ Many studies suggest an “ecological bias” when relying on solely on area-based deprivation indictors which may be because they overlook significant variations in individual deprivation living within the same community. However, the IMD score still demonstrates a notable correlation with these individual variations. Studies based on all-case mortality, cancer survival, and quality-adjusted life expectancy all found that area deprivation, with studies showing higher individual hospital admissions and poorer general health outcomes in areas with lower SES. This correlation suggests that IMD scores can still provide valuable insights into resource allocation and health interventions, particularly in assessing community health needs. Furthermore, the IMD score is a weighted score that incorporates various factors related to deprivation, including income, employment, education, health, crime, and environment. It collectively provides a holistic view of the socioeconomic context affecting communities. Both have a negative effect on outcome.^55
-57 Therefore, ecological deprivation scores, such as the IMD, should be used in conjunction with direct individual assessment tools to address the complexity of the multifactorial problem of deprivation and its impact on health outcomes.

This integrated approach ensures that the specific needs of populations are acknowledged and addressed effectively, ultimately contributing to more equitable health care strategies. Health inequalities are the systemic disparities in health that occur across SES, gender, ethnicity, sexual orientation, or other dimensions of society with varying access to material and nonmaterial resources.⁵⁸ Tackling health inequalities and equitable access to health care are pressing priorities for the UK’s NHS and other health care providers worldwide. In 2019, NHS England set a long-term plan to address health inequalities focusing on tackling access and health outcomes among the most deprived 20% of the population, launching an ambitious project, Core20PLUS5.⁵⁹ In addition, the implications of providing for the different health care needs of various social groups are that additional resources must be funded to deliver a comparable number of expected operation outcomes across deprived and nondeprived regions.⁶⁰ ACI represents a specialized treatment pathway, which requires a high degree of patient engagement, particularly in the rehabilitation phase, through regular attendance to physiotherapy sessions and adherence to specialized postoperative rehabilitation. Further research will help to identify the individual causes of inequality in post-ACI surgery outcomes in deprived areas, and whether these findings relate to, e.g., compliance with lengthy rehabilitation programs, access to practical physiotherapy sessions in the hospital or in the community. This level of detail is required to assist in tailoring more targeted interventions aiming to reduce ACI outcome inequalities associated with SES, which are suggested as the main findings of this study.

Conclusion

The study presented has demonstrated poorer functional outcomes following ACI surgery among patients from socioeconomically deprived areas. Moreover, ACI patients from higher unemployment level areas tended to have lower mean Lysholm scores at 12 months postoperatively. Our findings suggest that socioeconomic deprivation could be a confounding factor in functional outcomes following ACI surgery and, therefore, should be included as a relevant parameter in future predictive modeling studies on ACI outcome. Perhaps most importantly, the findings from this study have highlighted the importance of looking deeply into the effect of socioeconomic area deprivation on an individual level and whether its effect is related to different aspects of care pathways, e.g., compliance with specialist rehabilitation. This study additionally points to the importance of targeting health care resources to patients from deprived areas to improve surgical outcomes and reduce inequalities across health care settings.

Footnotes

Acknowledgements

The authors of this study would like to acknowledge the remarkable work of the late Professor James Richardson for establishing ACI treatment at the RJAH Orthopaedic Hospital; the majority of the cohort presented here were his patients. The authors are also grateful to Versus Arthritis (Grants 18480, 19429, 21156, and 20815), the Medical Research Council (MR/L010453/1 and MR/N02706X/1), and the Orthopaedic Institute Ltd in Oswestry (Sports Knee Fund) for their support of this work. Finally, we acknowledge the support of the National Institute for Health and Care Research (NIHR) Birmingham Biomedical Research Centre (BRC).

ORCID iDs

Salam T. Ismael

Helen S. McCarthy

Karina Wright

Sally Roberts

Jan Herman Kuiper

Ethical Considerations

The study titled “Socioeconomic area deprivation is related to poorer patient-reported outcomes following Autologous Chondrocyte Implantation (ACI)” is considered a service evaluation rather than a research study. The primary aim of this evaluation is to assess the relationship between socioeconomic area deprivation and a patient study titled “Socioeconomic Area Deprivation is Related to Poorer Patient-Reported Outcomes Following Autologous Chondrocyte Implantation (ACI)” was deemed to be a service evaluation rather than a research study. The primary objective of this study is to investigate the correlation between the level of deprivation and patients’ reported outcomes (PROMs) following ACI, in order to improve service delivery, identify trends, and inform clinical practice. The study utilized existing data collected as part of routine clinical care, focusing on the evaluation of health outcomes within a specific patient population. As the study did not involve any new interventions or research beyond the clinical care provided, and it did not involve randomization, experimental manipulation, or collection of new data outside of routine care, it was not considered to meet the definition of research as outlined by ethical review boards. The study was, therefore, categorized as a service evaluation aimed at improving the quality and delivery of existing services rather than testing a hypothesis or generating generalizable knowledge. Given that the data reviewed was part of regular clinical practice and patient care, and no additional risks, interventions, or changes in the standard of care were introduced, it was concluded that formal ethical approval by an Institutional Review Board (IRB) or Research Ethics Committee (REC) was not required. The study was conducted by the relevant guidelines, ensuring patient confidentiality and compliance with data protection regulations.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

References

Delamothe

. Universality, equity, and quality of care. BMJ. 2008;336(7656):1278-81.

Winkleby

Jatulis

Frank

Fortmann

. Socioeconomic status and health: how education, income, and occupation contribute to risk factors for cardiovascular disease. Am J Public Health. 1992;82(6):816-20.

Sutton

Gravelle

Morris

Leyland

Windmeijer

Dibben

, et al. Allocation of resources to English areas: individual and small area determinants of morbidity and use of healthcare resources. London: Department of Health in England; 2002.

Dixon

Le Grand

Henderson

Murray

Poteliakhoff

. Is the British national health service equitable? The evidence on socioeconomic differences in utilization. J Health Serv Res Policy. 2007;12(2):104-9.

Morris

Sutton

Gravelle

. Inequity and inequality in the use of health care in England: an empirical investigation. Soc Sci Med. 2005;60(6):1251-66.

NHS. Core20PLUS5 (adults)—an approach to reducing healthcare inequalities. https://www.england.nhs.uk/about/equality/equality-hub/national-healthcare-inequalities-improvement-programme/core20plus5/

Brittberg

Lindahl

Nilsson

Ohlsson

Isaksson

Peterson

. Treatment of deep cartilage defects in the knee with autologous chondrocyte transplantation. N Engl J Med. 1994;331(14):889-95.

Peterson

Minas

Brittberg

Nilsson

Sjögren-Jansson

Lindahl

. Two- to 9-year outcome after autologous chondrocyte transplantation of the knee. Clin Orthop Relat Res. 2000;(374):212-34.

McCarthy

McCall

Williams

Mennan

Dugard

Richardson

, et al. Magnetic resonance imaging parameters at 1 year correlate with clinical outcomes up to 17 years after autologous chondrocyte implantation. Orthop J Sports Med. 2018;6(8):2325967118788280.

10.

Pareek

Carey

Reardon

Peterson

Stuart

Krych

. Long-term outcomes after autologous chondrocyte implantation: a systematic review at mean follow-up of 11.4 years. Cartilage. 2016;7(4):298-308.

11.

Gikas

Bayliss

Bentley

Briggs

. An overview of autologous chondrocyte implantation. J Bone Joint Surg Br. 2009;91(8):997-1006.

12.

Ebert

Smith

Edwards

Hambly

Wood

Ackland

. Factors predictive of outcome 5 years after matrix-induced autologous chondrocyte implantation in the tibiofemoral joint. Am J Sports Med. 2013;41(6):1245-54.

13.

Hollander

Dickinson

Sims

Brun

Cortivo

Kon

, et al. Maturation of tissue engineered cartilage implanted in injured and osteoarthritic human knees. Tissue Eng. 2006;12(7):1787-98.

14.

Roberts

Hollander

Caterson

Menage

Richardson

. Matrix turnover in human cartilage repair tissue in autologous chondrocyte implantation. Arthritis Rheum. 2001;44(11):2586-98.

15.

Roberts

Menage

Sandell

Evans

Richardson

. Immunohistochemical study of collagen types I and II and procollagen IIA in human cartilage repair tissue following autologous chondrocyte implantation. Knee. 2009;16(5):398-404.

16.

Paatela

Vasara

Nurmi

Kautiainen

Jurvelin

Kiviranta

. Biomechanical changes of repair tissue after autologous chondrocyte implantation at long-term follow-up. Cartilage. 2021;13(1 Suppl):1085S-1091S.

17.

Campbell

Pineda

Harris

Flanigan

. Return to sport after articular cartilage repair in athletes’ knees: a systematic review. Arthroscopy. 2016;32(4):651-68.

18.

Galvin

Agrawal

Curry

. The impact of socioeconomic status on outcomes in orthopaedic surgery. J Bone Joint Surg Am. 2020;102(5):428-44.

19.

Barrack

Ruh

Chen

Lombardi

Jr Berend

Parvizi

, et al. Impact of socioeconomic factors on outcome of total knee arthroplasty. Clin Orthop Relat Res. 2014;472(1):86-97.

20.

Dugard

Kuiper

Parker

Roberts

Robinson

Harrison

, et al. Development of a tool to predict outcome of autologous chondrocyte implantation. Cartilage. 2017;8(2):119-30.

21.

Minas

Von Keudell

Bryant

Gomoll

. The John Insall award: a minimum 10-year outcome study of autologous chondrocyte implantation. Clin Orthop Relat Res. 2014;472(1):41-51.

22.

Hawker

Davis

French

Cibere

Jordan

March

, et al. Development and preliminary psychometric testing of a new OA pain measure—an OARSI/OMERACT initiative. Osteoarthritis Cartilage. 2008;16(4):409-14.

23.

Tegner

Lysholm

. Rating systems in the evaluation of knee ligament injuries. Clin Orthop Relat Res. 1985;(198):43-9.

24.

Smith

Richardson

Tennant

. Modification and validation of the Lysholm knee scale to assess articular cartilage damage. Osteoarthritis Cartilage. 2009;17(1):53-8.

25.

Communities UMoHa. English indices of deprivation 2019 [cited 2019]. Available from: https://imd-by-postcode.opendatacommunities.org/imd/2019.

26.

Enders

. Applied missing data analysis. New York: Guilford Press; 2022.

27.

Victoria Savalei

CFF

. Robust two-stage approach outperforms robust full information maximum likelihood with incomplete nonnormal data. Struct Equ Model. 2014;21(2):280-302.

28.

Leffondre

Abrahamowicz

Siemiatycki

Rachet

. Modeling smoking history: a comparison of different approaches. Am J Epidemiol. 2002;156(9):813-23.

29.

Kocher

Steadman

Briggs

Zurakowski

Sterett

Hawkins

. Determinants of patient satisfaction with outcome after anterior cruciate ligament reconstruction. J Bone Joint Surg Am. 2002;84(9):1560-72.

30.

Ziedas

Abed

Swantek

Cross

Chaides

Rahman

, et al. Social determinants of health influence access to care and outcomes in patients undergoing anterior cruciate ligament reconstruction: a systematic review. Arthroscopy. 2022;38(2):583-94.

31.

Yoon

Kim

Park

Yang

. Accessibility for rehabilitation therapy according to socioeconomic status in patients with stroke: a population-based retrospective cohort study. Brain Neurorehabil. 2023;16(2):e16.

32.

Aljabri

Bulkeley

Cusick

. Telerehabilitation in the Middle East North Africa region: a structured review. Int J Telerehabil. 2021;13(2):e6401.

33.

Rucinski

Cook

Crecelius

Stucky

Stannard

. Effects of compliance with procedure-specific postoperative rehabilitation protocols on initial outcomes after osteochondral and meniscal allograft transplantation in the knee. Orthop J Sports Med. 2019;7(11):2325967119884291.

34.

Gazendam

Zhu

Chang

Phillips

Bhandari

. Virtual reality rehabilitation following total knee arthroplasty: a systematic review and meta-analysis of randomized controlled trials. Knee Surg Sports Traumatol Arthrosc. 2022;30(8):2548-55.

35.

Truong

Mosewich

Holt

Miciak

Whittaker

. Psychological, social and contextual factors across recovery stages following a sport-related knee injury: a scoping review. Br J Sports Med. 2020;54(19):1149-56.

36.

Truong

Bekker

Whittaker

. Removing the training wheels: embracing the social, contextual and psychological in sports medicine. Br J Sports Med. 2021;55(9):466-7.

37.

Kivimäki

Batty

Pentti

Shipley

Sipilä

Nyberg

, et al. Association between socioeconomic status and the development of mental and physical health conditions in adulthood: a multi-cohort study. Lancet Public Health. 2020;5(3):e140-e149.

38.

Bhosale

Kuiper

Johnson

Harrison

Richardson

. Midterm to long-term longitudinal outcome of autologous chondrocyte implantation in the knee joint: a multilevel analysis. Am J Sports Med. 2009;37 Suppl 1:131S-138S.

39.

Niemeyer

Köstler

Salzmann

Lenz

Kreuz

Südkamp

. Autologous chondrocyte implantation for treatment of focal cartilage defects in patients age 40 years and older: a matched-pair analysis with 2-year follow-up. Am J Sports Med. 2010;38(12):2410-6.

40.

Rosenberger

Gomoll

Bryant

Minas

. Repair of large chondral defects of the knee with autologous chondrocyte implantation in patients 45 years or older. Am J Sports Med. 2008;36(12):2336-44.

41.

Gille

Schuseil

Wimmer

Gellissen

Schulz

Behrens

. Mid-term results of autologous matrix-induced chondrogenesis for treatment of focal cartilage defects in the knee. Knee Surg Sports Traumatol Arthrosc. 2010;18(11):1456-64.

42.

Gillinov

Fosam

Burroughs

Schneble

McLaughlin

Moran

, et al. Incidence, timing, and risk factors for 5-year revision surgery after autologous chondrocyte implantation in 533 patients. Am J Sports Med. 2022;50(11):2893-9.

43.

Behery

Harris

Karnes

Siston

Flanigan

. Factors influencing the outcome of autologous chondrocyte implantation: a systematic review. J Knee Surg. 2013;26(3):203-11.

44.

Jaiswal

Bentley

Carrington

Skinner

Briggs

. The adverse effect of elevated body mass index on outcome after autologous chondrocyte implantation. J Bone Joint Surg Br. 2012;94(10):1377-81.

45.

Birkenes

Furnes

Laastad Lygre

Solheim

Aaroen

Knutsen

, et al. The long-term risk of knee arthroplasty in patients with arthroscopically verified focal cartilage lesions: a linkage study with the Norwegian arthroplasty register, 1999 to 2020. J Bone Joint Surg Am. 2023;105(12):951-61.

46.

Wang

Shan

Kang

Yang

, et al. Association of acute postoperative pain and cigarette smoking with cerebrospinal fluid levels of beta-endorphin and substance P. Front Mol Neurosci. 2021;14:755799.

47.

Jaiswal

Macmull

Bentley

Carrington

Skinner

Briggs

. Does smoking influence outcome after autologous chondrocyte implantation?: a case-controlled study. J Bone Joint Surg Br. 2009;91(12):1575-8.

48.

Betz

Holzgruber

Simon

Uhlemann

Niemeyer

Müller

, et al. The effect of smoking on the outcome of matrix-based autologous chondrocyte implantation: data from the German cartilage registry. J Knee Surg. 2023;36(2):181-7.

49.

Endler

Ekman

Hellström

Möller

Gerdhem

. Minor effect of loss to follow-up on outcome interpretation in the Swedish spine register. Eur Spine J. 2020;29(2):213-20.

50.

Newman

. Missing data: five practical guidelines. Organ Res Methods. 2014;17(4):372-411.

51.

Ross

O’Rourke

Toland

MacDonald

Clement

Scott

CEH

. Loss to patient-reported outcome measure follow-up after hip arthroplasty and knee arthroplasty: patient satisfaction, associations with non-response, and maximizing returns. Bone Jt Open. 2022;3(4):275-83.

52.

Imam

Barke

Stafford

Parkin

Field

. Loss to follow-up after total hip replacement: a source of bias in patient reported outcome measures and registry datasets? Hip Int. 2014;24(5):465-72.

53.

Rijk

Kortlever

JTP

Bandell

DLJI

Zhang

Gallagher

Bozic

, et al. The impact of socioeconomic status and social deprivation on musculoskeletal limitations. J Orthop. 2020;22135-42.

54.

Ingleby

Belot

Atherton

Baker

Elliss-Brookes

Woods

. Assessment of the concordance between individual-level and area-level measures of socio-economic deprivation in a cancer patient cohort in England and Wales. BMJ Open. 2020;10(11):e041714.

55.

Ribeiro

Fraga

Severo

Kelly-Irving

Delpierre

Stringhini

, et al. Association of neighbourhood disadvantage and individual socioeconomic position with all-cause mortality: a longitudinal multicohort analysis. Lancet Public Health. 2022;7(5):e447-e457.

56.

Woods

Belot

Atherton

Ellis-Brookes

Baker

Ingleby

. Are deprivation-specific cancer survival patterns similar according to individual-based and area-based measures? A cohort study of patients diagnosed with five malignancies in England and Wales, 2008-2016. BMJ Open. 2022;12(6):e058411.

57.

Murphy

Hinde

Richardson

. Appropriate categorization of inequality to inform policy decisions: estimating distribution of lifetime health using alternative approaches to socioeconomic stratification. Value Health. 2024;27(1):26-34.

58.

Eikemo

Øversveen

. Social inequalities in health: challenges, knowledge gaps, key debates and the need for new data. Scand J Public Health. 2019;47(6):593-7.

59.

England

. National healthcare inequalities improvement programme 2019. Available from: https://www.england.nhs.uk/about/equality/equality-hub/national-healthcare-inequalities-improvement-programme/.

60.

Barnard

Grant

Hickey

Bridgewater

. Is social deprivation an independent predictor of outcomes following cardiac surgery? An analysis of 240 221 patients from a national registry. BMJ Open. 2015;5(6):e008287.