Evaluation of functional outcomes and patient satisfaction after unicompartmental knee arthroplasty: a retrospective analysis from 2011 to 2022

Full Text

INTRODUCTION

Knee osteoarthritis is the most common cause of knee pain and functional impairment in elderly patients. Due to population aging and global obesity epidemic, the incidence of knee osteoarthritis has increased significantly. Although total knee arthroplasty is considered the gold standard of the surgical treatment of patients with end-stage knee osteoarthritis, 10-20% of patients remain not fully satisfied with the treatment outcomes [1]. This has led to growing interest in alternative surgical approaches, particularly unicompartmental knee arthroplasty (UKA).

Among the advantages of partial knee arthroplasty compared with total knee arthro-plasty are lower surgical invasiveness and shorter operative time, minimal blood loss, fas-ter postoperative recovery, better knee function and subjective joint perception, lower rates of infectious and systemic complications, as well as technical simplicity of revision procedures when required [2, 3, 4].

Recent studies have confirmed the rationale for expanding the indications for UKA, including its use in post-traumatic knee osteoarthritis with anterior cruciate ligament deficiency, in middle-aged patients, in cases of spontaneous osteonecrosis of the femoral condyles, and even in patients with rheumatologic diseases in remission [5, 6, 7, 8].

Over the past decades, a decrease in revision rates following partial knee arthroplasty has been reported, along with the improvements in functional outcome scores and patient-reported satisfaction. These improvements are particularly evident when outcome measures without a ceiling effect — such as FJS-12 and OKS — are used, as opposed to traditional scoring systems (KSS, WOMAC), which are less sensitive in differentiating patients with favorable outcomes [9, 10, 11, 12, 13].

Unfortunately, in the Russian Federation, the share of UKA among all knee arthro-plasty procedures performed annually remains negligible. Even at the Vreden National Me-dical Research Center of Traumatology and Orthopedics, according to the data from the local arthroplasty registry, UKA has accounted for no more than 5.9% of knee replacement surgeries over the past decade [14], although for some surgeons at the center this share may reach 32%. Among the reasons for the limited popularity of partial knee arthroplasty, Russian orthopedic surgeons specializing in joint replacement note that in patients awaiting hospitalization, medial compartment osteoarthritis often progresses to total joint involvement, or that suitable candidates for UKA are relatively rare [15].

The present study is dedicated to address the existing gap in knowledge regarding residual pain and knee joint function in mid- and long-term follow-up after UKA.

The aim of the study — to evaluate functional outcomes and patient satisfaction following unicompartmental knee arthroplasty, as well as to determine the impact of demographic factors and obesity on clinical outcomes and subjective pain perception during various activities of daily living.

METHODS

Study design

The present study is a retrospective analysis of the outcomes of medial UKA performed between 2011 and 2022. Of the total 1.178 UKA procedures carried out during the study period, 33 opera-tions using a fixed tibial bearing were excluded, as well as 3 lateral compartment UKAs. In addi-tion, procedures performed in patients who had undergone any revision surgery of the knee joint for any reason were excluded (4 due to persistent pain, 2 due to component instability, 1 due to joint instability, and 1 due to insert dislocation). After the application of the inclusion criteria, 302 patients with primary medial UKA who could be contacted by phone and had complete data on patient satisfaction and functional outcomes with a minimum follow-up of 12 months were included in the study.

Patients

Women predominated in the study cohort, with 248 female patients (82.1%) and 54 male patients (17.9%), which corresponds to the typical sex distribution observed in knee osteoarthritis, with a higher prevalence among women [16]. The mean age at the time of sur-gery was 67.8±9.2 years (range, 42-89 years). The age distribution was as follows: 18-44 years (n = 1; 0.3%), 45-59 years (n = 54; 17.9%), 60-74 years (n = 191; 63.2%), and 75-90 years (n = 56; 18.5%). Due to the insufficient sample size in the 18-44-year age group, this category was excluded from the age-related analysis. Thus, the final sample for age-based analysis consisted of 301 observations. The mean body mass index (BMI) was 30.8±4.9 kg/m² (range, 18.7-46.2 kg/m²). Pati-ents were stratified into five categories according to the WHO classification: normal weight (18.5-24.9 kg/m²) — 31 patients (10.3%); over-weight (25.0-29.9 kg/m²) — 87 patients (28.8%); obesity class I (30.0-34.9 kg/m²) — 114 patients (37.7%); obesity class II (35.0-39.9 kg/m²) — 56 pa-tients (18.5%); and obesity class III (≥ 40 kg/m²) — 14 patients (4.6%).

Methods of clinical outcome assessment

Patient satisfaction was assessed using a 5-point Rasch scale ranging from 0 to 5, where 0 indicated complete dissatisfaction and 5 indicated maximal satisfaction [12]. In addition, a dichotomous variable, “high satisfaction”, was introduced and defined as a score of 4 or 5 to facilitate the interpretation of the results and enable categorical analysis.

Methods of pain assessment and analysis

A 5-point ordinal Likert scale was used to quantitatively assess subjective pain frequency, including the following categories: “never”, “almost never”, “occasionally”, “often”, and “almost always”. This scale was selected as an optimal tool for converting qualitative subjective assessments into quantitative measures, in accordance with contemporary approaches to pain evaluation.

To allow for statistical analysis and interpretation, each category was assigned a numerical value: “never” = 0, “almost never” = 1, “occasionally” = 2, “often” = 3, and “almost always” = 4. Numerical coding was applied for the calculation of composite scores and descrip-tive statistics. Given the ordinal nature of the scales and the results of normality testing, non-parametric methods were used for intergroup comparisons.

Based on the obtained numerical values, two composite indices were calculated.

The pain index (PI) is a weighted mean calculated using the following formula:

PI = (0×N₀ + 1×N₁ + 2×N₂ + 3×N₃ + 4×N₄) / / (N₀ + N₁ + N₂ + N₃ + N₄),

where N₀, N₁, N₂, N₃, N₄ represent the number of responses in the corresponding categories.

This index yields a single numerical value ranging from 0 to 4, where 0 indicates complete absence of pain and 4 indicates the maximum frequency of pain. The use of a weighted mean accounts for the distribution of responses across all categories and allows for comparative analysis between groups and over time.

Interpretation of the PI:

• 0.0-0.5 — excellent outcome (minimal pain);
• 0.5-1.5 — good outcome;
• 1.5-2.5 — fair outcome;
• 2.5-3.5 — poor outcome;
• 3.5-4.0 — very poor outcome (severe pain).

The pain-free percentage (PFP) reflects the relative proportion of pain-free status and is calculated using the following formula:

PFP = (N₀ + 0.75×N₁ + 0.5×N₂ + 0.25×N₃ + 0×N₄) / / (N₀ + N₁ + N₂ + N₃ + N₄) × 100%.

This index presents the results in a more intuitive percentage format, where 100% corresponds to complete absence of pain and 0% corresponds to constant pain. The weighting coefficients (0.75, 0.50, and 0.25) were selected to linearly reflect decreasing pain-free status, in accordance with recommendations for constructing composite pain assessment scales.

The use of two complementary indices was motivated by the need for a comprehensive eva-luation of pain and is consistent with the prin-ciple of data triangulation in medical statistics. The PI allows for more precise differentiation of symptom severity, whereas the PFP enhances the clarity of result presentation and facilitates interpretation in clinical practice.

Statistical analysis

Preliminary data analysis included the assessment for missing values and evaluation of distribution normality. The Shapiro-Wilk test was used to assess normality and demonstrated that the distribution of all analyzed variables was non-normal (p <  0.05).

Descriptive statistics included the mean, stan-dard deviation, median, and interquartile range. Given the non-normal distribution of the data and the ordinal nature of most scales, non-parametric methods were predominantly used for statistical analysis.

The comparisons of quantitative and ordinal variables between two independent groups (e.g., women vs men) were performed using the Mann-Whitney U test. For the comparisons of quantitative and ordinal variables across three or more independent groups (formed according to age categories, BMI categories, or follow-up periods), the Kruskal-Wallis test was applied. When a significant overall effect was detected, post-hoc analysis was conducted using the Mann-Whitney U test with Bonferroni correction.

Changes in variables within paired samples (e.g., preoperative vs postoperative measurements) were assessed using the Wilcoxon signed-rank test. The comparisons of proportions and categorical variables were performed using Pearson’s chi-square test (r×c contingency tables). When expected cell frequencies were less than 5, exact tests (Fisher’s exact test) were applied.

Associations between variables were assessed using Spearman’s rank correlation coefficient (ρ). Pearson’s correlation coefficient was not applied to empirical data. The only reported correlation coefficient is provided to illustrate the deterministic linear relationship between the PFP and the PI (PFP = 100×(1 − PI/4)). To quantify the magnitude of differences between groups, effect size was calculated using Cohen’s d and interpreted as follows: d <  0.2, negligible effect; 0.2 ≤ d <  0.5, small effect; 0.5 ≤ d <  0.8, medium effect; and d ≥ 0.8, large effect.

To assess the effect of follow-up period and other factors on the likelihood of high patient satisfaction, a logistic regression model was constructed with the calculation of odds ratios (OR) and 95% confidence intervals.

Statistical power analysis demonstrated that a sample size of 302 patients provided high power to detect clinically meaningful differences. Given the small effect size observed in the study (Cohen’s d = 0.24 for sex-related differences), the achieved statistical power was 0.991, which substantially exceeds the commonly accepted threshold of 0.8. Power calculations indicated that to detect a small effect (d = 0.2) with a power of 0.8 at a significance level of α = 0.05, a sample size of 787 patients would be required; for a medium effect (d = 0.5), 128 patients; and for a large effect (d = 0.8), 52 patients. Thus, the actual sample size (n = 302) was more than sufficient to detect medium and large effects and provided adequate power even for identifying small effects. This confirms the robustness of the obtained results and supports the validity of conclusions regarding the absence of statistically significant differences between the studied groups.

Parametric tests (t-test, ANOVA, and Pearson’s correlation) were not applied to the empirical data due to the non-normal distribution of the variables.

The threshold for statistical significance was set at α = 0.05 for all analyses. All calculations were performed using IBM SPSS Statistics version 26.0 (IBM, USA), as well as the Python programming language with the pandas, numpy, and scipy libraries for specialized analyses.

RESULTS

Sex

Data analysis revealed notable differences in pain dynamics between men and women. Women demonstrated a higher PI (1.30±0.12) than men (0.74±0.17). However, these differences were not statistically significant. The PFP in women remained relatively stable during the first two periods (70.3±6.8% and 74.8±6.9%, respectively) but decreased substantially to 62.8% in the 2019-2022 period.

In men, an opposite trend was observed: the PFP decreased from 100 to 69.8% and subsequently partially recovered to 73.3%.

When satisfaction scores were compared between men and women, the median score in both groups was 5.0 [4.0; 5.0], with no statistically significant difference (U = 5871.5; p = 0.104; rank-biserial correlation rrb≈0.12, small effect in favor of men). The share of patients with high satisfaction (scores of 4-5) was 87.0% in men and 77.8% in women; χ²(1) = 2.31; p = 0.128.

The overall composite PI (across 12 activity types) for all periods was 1.11, with a mean PFP of 72.2%. The PFP was derived from the PI using the formula PFP = 100 × (1 − PI/4). The relation-ship between these indices is deterministic and was therefore not subjected to separate statisti-cal testing.

Notably, the most pronounced differences between men and women were observed during the 2011-2014 period, when men reported no pain, whereas women already experienced moderate pain. However, the interpretation of these findings is limited by the small size of the male subgroup during this period. By the end of the observation period (2019-2022), sex-related differences in pain outcomes had substantially diminished, which may indicate the convergence of functional outcomes between men and women in the long-term follow-up after UKA.

Age

The analysis of age-related satisfaction after UKA demonstrated high levels of patient satisfaction across all age groups. The mean satisfaction score was 4.11±1.13 in the 45-59-year group, 4.27±1.21 in the 60-74-year group, and 4.45±0.89 in the 75-90-year group. The median satisfaction score was 4.5 in the 45-59-year group and 5.0 in both the 60-74- and 75-90-year groups, with an interquartile range of 1.0 in all groups, indicating clustering of scores at the upper end of the scale.

The Kruskal-Wallis test revealed no statisti-cally significant differences between age groups (H = 3.40; p = 0.183). The share of patients with high satisfaction (scores of 4-5) increased gradually with age: 75.9% in the 45-59-year group, 80.1% in the 60-74-year group, and 80.4% in the 75-90-year group. However, these differences did not reach statistical significance.

The analysis of temporal trends within each age group revealed an important finding: a statistically significant improvement in satisfaction over time was observed only in the oldest age group (75-90 years) (H = 7.16; p = 0.028). In this group, a pronounced positive trend was noted, with mean satisfaction scores increasing from 3.67±0.95 in the 2011-2014 period to 4.53±0.82 in 2015-2018 and 4.60±1.01 in 2019-2022. This finding is particularly rele-vant, as it demonstrates progressive improve-ment in outcomes among the oldest patients.

In the middle-aged group (60-74 years), an opposite, although not statistically significant, trend was observed, with a slight decrease in satisfaction from 4.67 in 2011-2014 to 4.18 in 2019-2022 (H = 2.48; p = 0.289). In the younger group (45-59 years), changes were non-linear and did not demonstrate a clear trend (H = 1.67; p = 0.434) (Table 1).

 

Table 1

Patient satisfaction after unicompartmental knee arthroplasty depending on age and follow-up period

Parameter	45-59 years (n = 54)	60-74 years (n = 191)	75-90 years (n = 56)	p-value
Satisfaction, scores
Mean ± SD	4.11±1.13	4.27±1.21	4.45±0.89	0.183a
Median [Q1; Q3]	4.5 [4.0; 5.0]	5.0 [4.0; 5.0]	5.0 [4.0; 5.0]
Range	0-5	0-5	1-5
High satisfactionb, n (%)	41 (75.9)	153 (80.1)	45 (80.4)	0.784c
Temporal dynamics
2011-2014
n	3	16	7
Mean ± SD	5.00±0.00	4.67±0.62	3.67±0.82
Median [Q1; Q3]	5.0 [5.0; 5.0]	5.0 [4.8; 5.0]	4.0 [3.0; 4.0]
2015-2018
n	11	95	40
Mean ± SD	3.73±1.49	4.29±1.18	4.53±0.85
Median [Q1; Q3]	4.0 [3.0; 5.0]	5.0 [4.0; 5.0]	5.0 [4.0; 5.0]
2019-2022
n	40	80	9
Mean ± SD	4.19±0.98	4.18±1.31	4.56±1.01
Median [Q1; Q3]	4.0 [4.0; 5.0]	5.0 [3.3; 5.0]	5.0 [4.0; 5.0]
p-value of trendd	0.434	0.289	0.028

^a — Kruskal-Wallis test for comparison between age groups; ^b — satisfaction 4-5 scores on a 5-point scale; ^c — Pearson’s chi-square test; ^d — Kruskal-Wallis test for trend analysis within an age group; SD — standard deviation. Statistically significant values are shown in bold.

 

Correlation analysis revealed no significant association between age and satisfaction (ρ = -0.032; p = 0.585), confirming the absence of a linear relationship between these parameters and indicating that UKA provides high patient satisfaction regardless of patients’ age.

Overall, age does not appear to be a limiting factor in achieving high satisfaction after UKA. Moreover, patients in the oldest age group demonstrated a significant improvement in out-comes over time, which may reflect advances in surgical technique and perioperative manage-ment of elderly patients.

Body mass index

The impact of BMI on satisfaction after UKA was analyzed in 302 patients. The Kruskal-Wallis test revealed statistically significant differences in satisfaction across BMI categories (H = 40.82; p <  0.0001). The highest level of satisfaction was observed in patients with normal body weight, with a mean satisfaction score of 4.35±1.20 and a median of 5.0. In patients with elevated BMI, satisfaction scores were significantly lower: 3.38±1.00 in the overweight group; 3.32±1.08 in obesity class I; 3.36±1.00 in obesity class II; and 3.21±1.12 in obesity class III. In all of these groups, the median satisfaction score was 4.0.

A decreasing trend in the share of patients with high satisfaction (scores of 4-5) was observed with increasing BMI. However, differences between BMI categories did not reach statistical significance (χ²(4) = 7.87; p = 0.096). In patients with normal body weight, the share of highly satisfied patients reached 87.1%, whereas in groups with elevated BMI it was substantially lower: 59.8% in the over-weight group, 63.2% in obesity class I, 64.3% in obesity class II, and 64.3% in obesity class III (Table 2).

 

Table 2

Patient satisfaction after unicompartmental knee arthroplasty depending on body mass index

BMI	Total number of patients, n (%)	Satisfaction, scores	High satisfactiona
		Median [Q1; Q3]	n (%)
18.5-24.9	31 (10.3)	5.0 [4.0; 5.0]	27 (87.1)
25.0-29.9	87 (28.8)	4.0 [3.0; 4.0]	52 (59.8)
30.0-34.9	114 (37.7)	4.0 [3.0; 4.0]	72 (63.2)
35.0-39.9	56 (18.5)	4.0 [3.0; 4.0]	36 (64.3)
≥ 40	14 (4.6)	4.0 [2.0; 4.0]	9 (64.3)
p-valueb			0.096c

^a — satisfaction 4-5 scores on a 5-point Rasch scale; ^b — Kruskal-Wallis test; ^c — Pearson’s chi-square test.

 

To assess the effect of BMI in relation to follow-up periods, the entire observation period was divided into three intervals: 2011-2014 (n = 19), 2015-2018 (n = 156), and 2019-2022 (n = 127). No statistically significant differences in overall satisfaction were identified between these time periods. Mean satisfaction scores were 3.37±0.96 in 2011-2014, 3.53±1.06 in 2015-2018, and 3.36±1.15 in 2019-2022. The median satisfaction score remained stable at 4.0 across all periods, with an interquartile range of 3.0 to 4.0.

Unfortunately, the limited sample size in the 2011-2014 period (n = 19) precluded reliable analysis of changes in BMI category distribution over time. Correlation analysis revealed no significant association between the year of surgery and satisfaction (ρ = -0.018; p = 0.754), indicating the stability of UKA out-comes over the 12-year observation period regardless of patients’ BMI.

It is important to note that the absence of a temporal trend in outcomes may reflect a standardized approach to patient selection and surgical technique throughout the study period. This also supports the conclusion that the observed association between BMI and satisfaction represents a stable characteristic independent of the timing of surgery.

Pain outcomes during activities

A detailed analysis of pain during 12 different activities of daily living assessed using the FJS-12 questionnaire revealed substantial differences in pain perception between men and women, as well as specific patterns of functional limitation after UKA (Table 3).

 

Table 3

Pain index and pain-free percentage during different activities of daily living

Type of activity	Women (n = 248)		Men (n = 54)		∆ (W-M)
	Pain index	Pain-free percentage	Pain index	Pain-free percentage	Pain index
The most painful activities
Rising from a low object	1.72	56.9	1.32	67.0	+0.40
Climbing stairs	1.46	63.4	1.23	69.3	+0.23
Walking on uneven surfaces	1.37	65.7	1.12	72.1	+0.25
Household chores/gardening	1.33	66.7	1.22	69.4	+0.11
Sports/physical activity	1.32	67.0	0.96	76.0	+0.36
Prolonged standing	1.30	67.6	1.31	67.2	-0.01
Moderately painful activities
Walking >  15 minutes	1.26	68.5	1.22	69.4	+0.04
Sitting >  1 h	0.99	75.2	0.86	78.5	+0.13
Hiking/walking	0.97	75.8	0.83	79.3	+0.14
The least painful activities
Resting in bed	0.90	77.4	0.82	79.4	+0.08
Bathing/showering	0.89	77.6	0.46	88.5	+0.43**
Driving a car	0.80	79.9	0.77	80.6	+0.03

** p < 0.01 when compared between men and women.

 

The composite PI (across 12 activities) was 1.18±0.28 in women and 1.02±0.27 in men. The mean PFP was 70.3±6.8% in women and 74.8±6.9% in men. Although these differences did not reach statistical significance, a consistent trend toward higher pain levels in women was observed across most activity domains.

Sex-specific pain features

Analysis revealed characteristic sex differences in pain patterns. In women, the most problematic activities were those requiring substantial knee loading and muscular strength: rising from a low object (PI = 1.72), climbing stairs (PI = 1.46), and walking on uneven surfaces (PI = 1.37). In men, the highest pain levels were observed when rising from a low object (PI = 1.32) and du-ring prolonged standing (PI = 1.31).

The most substantial sex differences were noted during bathing/showering (∆ = 0.43), rising from a low object (∆ = 0.40), and sports/physical activity (∆ = 0.36), where women experienced substantially greater discomfort. The only activity in which men reported more pain was prolonged standing (∆ = -0.01).

Temporal dynamics of pain

In women, pain exhibited a U-shaped trend over time: after improvement in the 2015-2018 period (PI = 1.01), a significant deterioration occurred in 2019-2022 (PI = 1.49; p <  0.001). In men, a sharp transition was observed from complete absence of pain in 2011-2014 to moderate pain in the subsequent periods (Table 4).

 

Table 4

Temporal dynamics of pain syndrome

Parameter	2011-2014	2015-2018	2019-2022	p-valueа
Women
Pain index	1.05±0.31	1.01±0.26	1.49±0.18**	<  0.001
Pain-free percentage	73.9±7.7	74.7±6.5	62.8±4.5**	<  0.001
Number of patientsb	22	141	115
Men
Pain index	0.00±0.00	1.21±0.19†	1.07±0.17†	<  0.001
Pain-free percentage	100.0±0.0	69.8±4.8†	73.3±4.3†	<  0.001
Number of patientsb	4	31	31

^a — Kruskal-Wallis test for comparison between time periods; ^b — number of patients in the period; ** p <  0.01 when compared with the period 2015-2018; † — p <  0.001 when compared with the period 2011-2014.

 

Based on this analysis, three functional clus-ters of daily physical activities were identified:

1. complex activities (mean PI: women 1.44; men 1.19):

• rising from a low object,
• climbing stairs,
• walking on uneven surfaces,
• prolonged standing;

2. moderate-intensity activities (mean PI: women 1.14; men 1.03):

• household chores,
• walking > 15 minutes,
• hiking;

3. low-intensity activities (mean PI: women 0.90; men 0.73):

• resting in bed,
• bathing/showering,
• driving a car,
• sitting on a chair.

Despite the presence of residual pain in certain activities, the overall PFP remained relatively high (>  70% for most activity types), indicating the effectiveness of UKA in restoring patients’ daily functional mobility.

DISCUSSION

This study represents one of the largest retro-spective analyses of functional outcomes and pa-tient satisfaction following UKA in the Russian population.

The use of the Likert scale to assess pain across different activities, combined with the calculation of composite indicators (PI and PFP), represents an innovative approach to evaluating functional outcomes. Since PFP is linearly derived from PI (PFP = 100×(1 − PI/4)), a predetermined linear relationship exists between them. We consider PI and PFP as complementary representations of a single construct rather than as the result of a statistical correlation. This approach enables to overcome the limitations of traditional pain assessment scales and complements the FJS-12 scale [17, 18]. Previous studies have demonstrated that the FJS-12 is superior in detecting the “forgotten joint” phenomenon compared to conventional scales [19, 20]. The approach used in this study allows for the detailed assessment of specific aspects of pain in daily life, which is particularly important for patients after UKA, who typically achieve high functional outcomes.

The results demonstrate the high effectiveness of UKA and highlight several key aspects that warrant further discussion.

First, our study confirmed the overall high patient satisfaction after UKA (92.1% of patients reporting high satisfaction, median 5.0 points), which is comparable to the data reported by other authors [21, 22, 23, 24]. This finding is particularly relevant given that traditional assessment scales may exhibit a “ceiling effect” in patients after UKA [19, 25].

Our results also demonstrate a substantially higher level of satisfaction with UKA compared to outcomes after total knee arthroplasty. E.C. Rodriguez-Merchan, in a systematic review, reports that dissatisfaction following primary total knee arthroplasty ranges from 15 to 25% [26].

A meta-analysis by K. Xia et al., which included 13 randomized controlled trials with 683 patients after UKA and 683 patients after total knee arthroplasty, found no statistically significant differences in patient satisfaction between the two procedures (RR: 1.24; 95% CI: 0.92-1.66; p = 0.15) [27]. It is important to note, however, that this analysis included only patients with isolated medial femorotibial compartment osteoarthritis, representing the optimal population for UKA.

The trend observed in our study toward higher satisfaction among men (87.0%) compared to women (77.8%), although not statistically significant, aligns with the findings from the current literature. E.C. Rodriguez-Merchan identifies female sex as one of the main preoperative factors negatively affecting satisfaction after knee arthroplasty [26].

A systematic review by S. Shetty et al. confirms that female sex is associated with lower quality of life after knee arthroplasty. In four studies included in this review, women exhibited lower quality of life both preoperatively and postoperatively [28].

Our findings further support the well-known advantages of UKA, including reduced surgical trauma, faster recovery, and better postoperative functional outcomes. The high level of satisfaction observed across all age groups indicates that age should not be considered a limiting factor for UKA [29]. This aligns with the contemporary trend toward expanding indications for this procedure [30].

Notably, the observed improvement in satisfaction among the older age group (75-90 years) over time may reflect advancements in perioperative management as well as lower functional demands on the operated knee in this patient population [31].

Another key finding of our study is the identification of sex differences in pain percep-tion after UKA. Women reported a higher PI (1.3 in women vs 0.74 in men) while maintaining a high level of overall satisfaction. Importantly, these sex differences in pain did not reach statistical significance, which is consistent with other studies reporting no significant differences in long-term outcomes between men and women after UKA [32, 33].

Of particular interest is the U-shaped pattern of pain observed in women, with minimal levels recorded during 2013-2018, followed by a subsequent worsening in 2019-2022. A potential explanation for this phenomenon may be the impact of the COVID-19 pandemic on the accessibility and quality of postoperative rehabilitation.

The most problematic activities for women were rising from a low object, climbing stairs, and walking on uneven surfaces, all of which require functional stability and muscular strength. These differences highlight the need for a sex-specific approach to rehabilitation, with targeted training of these particular activities.

The absence of a statistically significant effect of BMI on patient satisfaction represents an important clinical observation. Although patients with morbid obesity (BMI ≥ 40 kg/m²) tended to report lower satisfaction (64.3% vs 87.1% in patients with normal weight), these differences did not reach statistical significance.

These findings partially contrast with some studies reporting a negative impact of obesity on UKA outcomes [34]. However, a meta-analysis by J. Lua et al. demonstrated that obesity should not be considered a relative contraindication for UKA, which is consistent with our results [35]. Notably, patients with obesity showed substantial improvements in functional outcomes after UKA [36]. On the other hand, our results partially align with a systematic review by S. Shetty et al., which revealed conflicting evidence regarding the influence of BMI on quality of life after total knee arthroplasty [28]. In 10 of 29 studies included, no association between BMI and postoperative outcomes was observed, while other studies reported both positive and negative effects of elevated BMI.

Although our study did not include a direct comparison of outcomes, a large meta-analysis conducted by K. Song et al. demonstrated that UKA is associated with better quality of life compared to total knee arthroplasty, regardless of patient BMI [37]. In their study, which included 64.585 patients, those undergoing UKA had significantly better EQ-5D scores (MD = -0.04; 95% CI: -0.05 … -0.02) and improved functional outcomes according to the Oxford Knee Score (MD = -1.29; 95% CI: -1.86 … -0.72) compared with total knee arthroplasty.

Interestingly, a study by K. Giesinger et al., conducted on a cohort of 1.565 total knee arthroplasty patients, reported the opposite trend: patients with class I-II obesity de-monstrated greater improvement in WOMAC functional scores compared to patients with normal weight [38]. The authors attributed this finding to the fact that patients with higher BMI had worse preoperative scores, thus possessing greater potential for postoperative improvement.

A systematic review by J. Li et al., which included 32 studies with 18.792 patients, identified 10 factors strongly associated with the development of chronic pain following total knee arthroplasty, among which BMI was a key factor [39]. The authors noted that elevated BMI, together with age, comorbidities, preoperative pain, and psychological factors, is a significant predictor of adverse outcomes. However, in our study of UKA, no such pattern was observed, suggesting potential differences in the pathophysiology of pain between total and partial knee arthroplasty.

The results of our study have important practical implications for clinical practice. In many centers, BMI is considered a limiting factor for total knee arthroplasty (usually 40 kg/m²). Our data suggest that such restrictions are not justified for UKA. This finding is particularly relevant given that obesity is a major risk factor for the development and progression of knee osteoarthritis, ultimately leading to the need for surgical intervention.

However, it should be noted that our study did not assess the incidence of complications, which, according to the literature, may be higher in patients with obesity. Therefore, the decision to perform UKA in patients with elevated BMI should be individualized, taking into account all risk factors and the potential benefits of the procedure.

Data from a large systematic review and meta-analysis by M. Poursalehian et al. demonstrate that UKA provides a more natural joint sensation compared to total knee arthroplasty [40]. Their analysis of 19 studies (4.500 knees) revealed a statistically significant difference in FJS-12 sco-res in favor of UKA (MD = 8.45; 95% CI: 6.03-10.88; p <  0.001), indicating lower patient awareness of the presence of an artificial joint.

These findings by M. Poursalehian et al. are corroborated by a systematic review by H.A. Wilson et al., which included 60 studies and demonstrated significant advantages of UKA across several key outcome measures [41]. Specifically, patients undergoing UKA had sta-tistically significantly shorter hospital stays across all three study groups: randomized controlled trials (MD = -1.20 days; 95% CI: -1.67 … -0.73), national registries (MD = -1.43 days; 95% CI: -1.53 … -1.33), and cohort studies (MD = -1.73 days; 95% CI: -2.30 … -1.16).

Clinical implications and personalized approach

The observed sex differences in pain per-ception have important clinical implications. A personalized rehabilitation approach for women, particularly regarding activities that require functional stability, should be considered when planning postoperative management. This aligns with the evidence suggesting that targeted, individualized rehabilitation can improve patient satisfaction [42].

Consistently high satisfaction levels across all age groups in our study confirm the universality of UKA. Recent studies indicate that patients over 80 years of age can successfully achieve favorable functional outcomes following UKA [31].

Study limitations

The retrospective study design and uneven distribution of patients across subgroups, particularly the predominance of women (82.1%), may have influenced the statistical power of the analysis.

Future research directions

To gain a more comprehensive understanding of factors influencing UKA outcomes, pros-pective studies incorporating objective func-tional assessments, such as instrumented gait analysis, muscle strength evaluation, and proprioception testing, are warranted. Of particular interest is investigating the impact of robotic-assisted surgery on UKA outcomes, as data from M. Poursalehian et al. suggest a trend toward superior FJS-12 outco-mes with robotic UKA compared to conventional techniques (MD = 10.12 vs 7.72).

Further research is also needed to develop personalized rehabilitation protocols that account for identified gender differences in pain and functional limitations. Long-term outcomes of UKA in patients with varying BMI, including implant survival and quality of life, remain an important area for future study.

CONCLUSION

The results of this study confirm the high effectiveness of medial unicompartmental knee arthroplasty in achieving patient satisfaction and restoring functional mobility, regardless of sex, age, or body mass index. The identified gender differences in pain patterns highlight the need for a differentiated approach to rehabilitation and postoperative management.

Taken together with the findings from contemporary meta-analyses, these results suggest that unicompartmental knee arthroplasty can be considered a preferred treatment option for patients with isolated medial knee osteoarthritis, providing a more natural joint feeling and superior functional outcomes compared to total knee arthroplasty. However, the final choice of treatment should be based on an individualized assessment of each patient, taking into account all risk factors, patient expectations, and the surgeon’s experience.

DISCLAIMERS

Author contribution

All authors made equal contributions to the study and the publication.

All authors have read and approved the final version of the manuscript of the article. All authors agree to bear responsibility for all aspects of the study to ensure proper consideration and resolution of all possible issues related to the correctness and reliability of any part of the work.

Funding source. This study was not supported by any external sources of funding.

Disclosure competing interests. The authors declare that they have no competing interests.

Ethics approval. Not applicable.

Consent for publication. Not required.

Use of artificial intelligence. No generative artificial intelligence technologies were used in the preparation of this manuscript.

About the authors

Alexey S. Fil

Vreden National Medical Research Center of Traumatology and Orthopedics

Author for correspondence.
Email: filalekse@yandex.ru
ORCID iD: 0000-0002-8994-0984
SPIN-code: 1413-8802

Cand. Sci. (Med.)

Russian Federation, St. Petersburg

Alexander P. Antipov

Vreden National Medical Research Center of Traumatology and Orthopedics

Email: a.p.antipov@ya.ru
ORCID iD: 0000-0002-9004-5952
SPIN-code: 8997-8235
Russian Federation, St. Petersburg

Taras A. Kuliaba

Vreden National Medical Research Center of Traumatology and Orthopedics

Email: taraskuliaba@mail.ru
ORCID iD: 0000-0003-3175-4756
SPIN-code: 1928-5286

Dr. Sci. (Med.)

Russian Federation, St. Petersburg

Nikolai N. Kornilov

Vreden National Medical Research Center of Traumatology and Orthopedics

Email: drkornilov@hotmail.com
ORCID iD: 0000-0001-6905-7900
SPIN-code: 9513-1362

Dr. Sci. (Med.), Professor

Russian Federation, St. Petersburg

References

Supplementary files