Sage Journals: Discover world-class research

Abstract

Aims

This study aims to evaluate the long-term durability of a unicompartmental knee arthroplasty (UKA) implant using a fixed-bearing all-polyethylene tibial tray with a Cobalt-Chromium (Co-Cr) femoral condyle, stratified by patient age. It also aims to compare implant survivorship between inlay and onlay tibial implantation techniques.

Methods

Data from the New Zealand Joint Registry (NZJR) was used to evaluate survival of Smith & Nephew Genesis II^® UKA performed between February 2000 and December 2011 in which two different tibial implant techniques (inlay and onlay) were used. The primary endpoint was revision surgery.

Results

At a minimum follow-up of 10 years, 57 revisions were performed at an average of 6.6 years from index surgery. The revision rate per 100 component years was 1.29 (CI: 0.97 – 1.66), with a significant difference between the over-75 years cohort compared with the under-65 cohort, in favour of the older cohort (p < 0.05). There was a significant difference in survival between inlay and onlay techniques, in favour of the inlay technique (p < 0.05).

Conclusion

Fixed-bearing all-polyethylene tibial UKA show acceptable outcomes at an average 12-year follow-up, especially in elderly patients and use of inlay tibial tray implantation technique.

Keywords

all-polyethylene tibia unicompartmental knee arthroplasty inlay onlay

Introduction

Unicompartmental knee arthroplasty (UKA) remains a valuable treatment option for knee arthritis despite the widespread success of total knee arthroplasty (TKA). Registry data suggest that UKA accounts for approximately 10% of knee arthroplasty procedures globally.^1,2 Controversy persists in the literature regarding decision-making in patient selection for UKA, including indications, prostheses and outcomes.^1,3–5 Proponents of UKA cite lower peri-operative morbidity, faster post-operative recovery, superior range of motion, better functional outcomes and preservation of bone stock.^3,6–10

The all-polyethylene tibial component has declined in global use due to concerns of early failure and limited perceived benefits over alternatives.^11,12 Nevertheless, its advantages of reduced cost, bone preservation and elimination of backside wear, remain relevant, particularly in select patient populations such as the elderly or resource-constrained settings.^11,13–15 There is a paucity of long-term outcome data specific to all-polyethylene UKA in large cohorts.^14,16

An underexplored aspect is the impact of tibial tray preparation technique, inlay versus onlay, on implant survivorship. Although biomechanical studies suggest differential stress distributions, the clinical significance of this in all-polyethylene implants is poorly defined.^11,12,16

The aim of this study was to address this knowledge gap by evaluating long-term outcomes of an UKA prosthesis with a fixed-bearing, all-polyethylene tibial tray articulating with a Co-Cr femoral runner related to age at index surgery and tibial tray implantation technique (inlay vs onlay).

Methods

The New Zealand Joint Registry (NZJR), with a capture rate of over 95%, was used to identify all Genesis II^® (Smith and Nephew) UKA over the active lifetime of this implant in New Zealand between 2000 and 2011.¹⁷ This implant was introduced into the New Zealand market in 2000 following withdrawal of the Mod III (Marmor) prosthesis and supplanted by a different implant in 2012. Follow-up data were analysed in 2025 to assess survivorship beyond 10 years. The NZJR has National ethical approval granted in 1998, with informed consent obtained from individuals undergoing arthroplasty procedures for collection and use of their data in research.

Demographic data and tibial tray implantation technique were collected. Inlay components were recessed into a carved pocket on the tibial surface and therefore relied more on subchondral bone, whereas onlay components were cemented on top of the flat tibial cut and therefore relied on cortical and subchondral bone (Figure 1). Interval to and reasons for revision surgery were analysed. Age at index operation was stratified into four bands (<55 years, 55–64 years, 65–75 years, ≥75 years) for sub-group analysis of implant survival.

Figure 1.

Inlay (a) and onlay (b) tibial implantation techniques.

Statistical analysis was performed using SPSS (IBM, USA). Comparisons between groups were performed using chi-square or Fisher’s exact tests, ANOVA and Kaplan-Meier survivorship analysis with log-rank tests. Cox proportional hazards regression was used to estimate the hazards ratio and 95% confidence interval comparing inlay and onlay groups. A p-value of 0.05 or less was considered statistically significant.

Revision rates in the NZJR were provided in the form of rate per 100 component years (RR/100CY). This was derived by dividing the number of implants revised by the observed component years, multiplied by 100. This allowed consideration of the number of years of post-operative follow-up in calculation of the revision rate. Implant survival was censored by mortality.

Results

Between February 2000 and December 2011, 350 Genesis II^® UKA procedures were performed by 19 surgeons. 91% were medial compartment replacements. Osteoarthritis was the main indication in 98% of cases, with an average patient age of 67 years. There were 164 (47%) inlay and 186 (53%) onlay tibial preparations. Patient demographics were comparable between inlay and onlay groups (Table 1). Significant differences between inlay and onlay cohorts were seen in primary surgeon (consultant vs trainee) and lateral compartment replacement (p < 0.01). In New Zealand, consultant supervision is mandatory for all trainee-led procedures.

Table 1.

Study demographics.

Study demographics	Inlay	Onlay
Frequency	46.9% (164/350)	53.1% (186/350)
Sex
Men	45.7% (75/164)	55.4% (103/186)	p = 0.072
Women	54.3% (89/164)	44.6% (83/186)
Average age (years)	68.2	66.8	p = 0.159
Age (frequency)
<55	9.1% (15/164)	7.5% (14/186)
55–64	33.5% (55/164)	40.9% (76/186)
65–74	31.1% (51/164)	29.6% (55/186)
≥75	26.2% (43/164)	22.0% (41/186)
Side
Left	51.2% (84/164)	52.2% (97/186)	p = 0.915
Right	48.8% (80/164)	47.8% (89/186)
Previous surgery
Yes	29.9% (49/164)	21.5% (40/186)	p = 0.085
No	70.1% (115/164)	78.5% (146/186)
Diagnosis
Osteoarthritis	98.8% (162/164)	97.3% (181/186)	p = 0.455
Other inflammatory	0	0.01% (1/186)	p = 1.00
Tumour	0.01% (1/164)	0	p = 0.469
Ligament disruption	0.01% (1/164)	0.01% (1/186)	p = 0.344
Compartment
Medial	96.3% (158/164)	86.6% (161/186)
Lateral	3.7% (6/164)	13.4% (25/186)	p < 0.001
6-months oxford knee score	38.36	36.04	p = 0.023
Average follow-up (years)	12.2	13
Primary surgeon (consultant)	98.8% (162/164)	84.9% (158/186)	p = 0.001

Mean follow-up was 12.6 years and average time to revision was 6.6 years (Table 1 and 2). 57 revisions occurred (1.29 RR/100CY; 95% CI: 0.97–1.66). Revision rate was highest in the 55–64 age cohort (1.96 RR/100CY; 95% CI: 1.35–2.70) and lowest in the ≥75 age cohort (0.35 RR/100CY; 95% CI: 0.07–1.02) (Table 2). The revision rate for the ≥75 age cohort was significantly lower than all other age cohorts (Table 2 and 3). These findings were also reflected in Kaplan-Meier survivorship analyses (Figure 2).

Table 2.

Revision rates between inlay and onlay cohorts.

Revisions	Inlay	Onlay	p-value
Frequency	11.0% (18/164)	21.0% (39/186)
Revision rate/100 component years	0.9	1.61	0.023
Revision rate/100 component years	95% CI: 0.53–1.43	95% CI: 1.13–2.18
Mean time to revision (Years)	8.3	5.7
Age	Rate/100-component-years
<55	0.44	2.37	0.079
<55	95% CI: 0.01–2.45	95% CI: 0.77–5.54
55–64	1.83	2.05	0.554
55–64	95% CI: 0.92–3.04	95% CI: 1.29–3.11
65–74	0.62	1.29	0.181
65–74	95% CI: 0.17–1.60	95% CI: 0.59–2.45
≥75	0	0.68	0.088
≥75	95% CI: 0–0.89	95% CI: 0.14–1.99
Revision for pain	6/164	19/186	0.027
Revision for pain	0.30 (95% CI: 0.11–0.65)	0.78 (95% CI: 0.47–1.23)
Revision for tibial component loosening	4/164	12/186	0.066
Revision for tibial component loosening	0.20 (95% CI: 0.05–0.51)	0.50 (95% CI: 0.26–0.87)

Table 3.

Implant survivorship by age and tibial preparation technique.

Survivorship (%)
Age	5-years	10-years	15-years
<55	93.10	93.10	82.76
95% CI	0.83–1.0	0.83–1.0	0.69–0.97
55–64	88.37	78.81	72.46
95% CI	0.82–0.94	0.72–0.86	0.65–0.80
65–74	94.30	88.16	86.95
95% CI	0.90–0.99	0.82–0.94	0.80–0.93
≥75	97.54	96.00	96.00
95% CI	0.94–1.00	0.92–1.00	0.92–1.00

Tibial preparation
Onlay	88.55	82.78	78.37
95% CI	0.84–0.93	0.77–0.88	0.72–0.85
Inlay	97.44	91.72	85.87
95% CI	0.95–1.00	0.87-0.96	0.80–0.92

Figure 2.

Cumulative Kaplan-Meier survivorship for 350 smith & nephew genesis II unicompartmental knee replacements stratified by age.

The inlay cohort had significantly lower revision rate (0.90 RR/100CY; 95% CI: 0.53–1.43) compared to the onlay cohort (1.61 RR/100CY; 95% CI: 1.13–2.18) (Hazards Ratio = 0.53, 95% CI: 0.30–0.92, p = 0.023) (Table 4). Time to revision on average, was longer in the inlay group (8.3 vs 5.7 years). The predominant reason for revision across the entire cohort was pain (43.9%) (Table 5). Revision due to pain was significantly higher in the onlay cohort (p = 0.027) (Table 2). Loosening of the tibial component was the second most common reason for revision (28.1%), with the incidence of this being 2.5 times greater in the onlay cohort than the inlay cohort (Table 2).

Table 4.

Revision rate per 100 component years for inlay and onlay cohorts.

Revision rate/100 component years
Overall UKA	1.29 (0.97–1.66)
Inlay	0.90 (0.53–1.43)
Onlay	1.61 (1.13–2.18)	p = 0.023
Hazards ratio = 0.53, 95% CI: 0.30–0.92

Table 5.

Reason for revision.

Reason for revision	Frequency	Percentage (%)
Pain	25	43.9
Loosening tibia	16	28.1
Other	11	19.3
Fracture tibia	3	5.3
Loosening femoral component	1	1.8
Deep infection	1	1.8
Total	57	100

Discussion

UKA remains a well-established and effective procedure in appropriately selected individuals with isolated end-stage compartmental knee arthritis. Compared to TKA, UKA is associated with preservation of cruciate ligaments and unaffected compartments, shorter operative time, reduced blood loss, higher patient reported outcome scores and lower risk of serious medical complications including venous thromboembolic phenomena.^6–10 Despite these advantages, concerns persist regarding UKA survival relative to TKA, as well as unresolved debates surrounding different UKA design concepts such as fixed versus mobile tibial bearings, metal-backed versus all-polyethylene tibial components and inlay versus onlay implantation of all-polyethylene tibial trays.^1,3,5

UKA survivorship and age

The revision rate in our cohort (1.29 per 100 component years) was higher than that of TKA in the NZJR (0.44 per 100 component years), consistent with prior reports showing UKA carries up to a four-fold increased revision risk compared to TKA.^2,18–23 Although UKA has a higher revision rate, conversion to TKA is generally technically less demanding than revision TKA, which may lower the surgical threshold for UKA revision. Outcomes after UKA to TKA revision remain mixed, with age at index surgery being a critical factor influencing survivorship.^22,24–27

A 15-years prospective randomized controlled trial by Newman et al. demonstrated superior survival in UKA compared to TKA in individuals over 65 years²⁸ Conversely, Pearse et al. found poorer outcomes in individuals under 65 undergoing revision UKA to TKA.²² These findings align with recommendations by the European Society of Sports Traumatology, Knee Surgery and Arthroscopy (ESSKA) UKA Manual, which supports UKA as a cost-effective and safe option for elderly patients.¹⁵

In our study, individuals aged 75 years and older had 15-years survivorship of 96%, consistent with literature reporting excellent outcomes in older cohorts (Table 3 and Figure 2). For instance, Bruce et al. reported a 10-years survivorship of 92.4% in individuals over 70, while NZJR data show highest UKA survival (94.8% at 10 years) in the 81–85 age group, with a progressive decline thereafter.^14,20

While these data support the durability of UKA in older populations, they must be interpreted with caution. The apparent survivorship advantage may be partly attributed to selection bias, as elderly patients typically have lower functional demands and higher competing mortality risks. These factors can reduce the likelihood of revision, thereby inflating survivorship estimates independent of implant performance.

UKA revision indications

In prior studies, aseptic loosening and osteoarthritis progression were the most common failure modes for UKA.^29,30 In contrast, our registry-based study identified “pain” as the leading reason for revision (43.9%), with osteoarthritis progression contributing less frequently (Table 5). This pattern is consistent with registry data reporting revision for pain in 48.8% of UKA cases and a higher incidence of unexplained pain after UKA compared to TKA (23% vs 9%, p < 0.001).^21,22

However, the prominence of pain as a revision indication represents a critical limitation when interpreting registry-based outcomes. Pain is inherently subjective and susceptible to variability in clinical interpretation, reporting thresholds, and coding practices. Chen et al. recently highlighted inaccuracies in the reporting of NZJR revision data, noting over-reporting of unexplained pain and under-reporting of objective causes for revision such as progression of osteoarthritis.³¹ These findings indicate that registry-reported revision indications may not reliably reflect true implant failure modes.

Accordingly, the high proportion of revisions attributed to pain in our study should be interpreted cautiously. Rather than indicating a mechanical failure mechanism, this likely reflects limitations in data collection and reporting compliance. This underscores the need for improved standardisation and validation of revision coding in arthroplasty registries to enhance the accuracy of outcome assessments and clinical interpretability of registry-derived outcomes.

Inlay versus onlay tibial implantation

Tibial component loosening is a growing cause of UKA failure.^28,30,32,33 Biomechanical studies suggest that inlay and onlay preparations affect stress distribution at the bone-implant interface. With an inlay preparation, a recess is created within the tibial plateau allowing the implant to fit into the prepared bone defect, which preserves bone around the outer edge of the tibia and captures the tibial tray. In comparison, an onlay preparation involves bone resection across the top of the tibial plateau of the affected compartment, typically for placement of a metal-backed plastic bearing, which is in theory, to be supported by the underlying cortical bone.¹² Walker et al. undertook finite element analysis based on computed tomography scans and found that inlay designs produce significantly higher peak stress and strain at the tibial surface compared to onlay components.¹²

Despite theoretical concerns of the influence of implant material on adaptive bone remodeling of the proximal tibia, clinical comparisons between inlay and onlay preparation techniques in all-polyethylene tibial components remain limited. In a Level III systematic review, Costa et al. found no significant difference between all-polyethylene and metal-backed designs.¹¹ Similarly, Lee et al. reported equivalent outcomes between metal-backed and all-polyethylene tibial components in a fixed-bearing UKA over 10 years¹³ All-polyethylene fixed-bearing tibial designs reduce revision risk due to bearing dislocation, backside wear observed in metal-backed tibial trays and depending on tibial preparation technique, may preserve bone stock by reducing bone cut from the tibia. Polyethylene wear has less frequently been cited as a cause of failure after UKA.¹⁶

We found a significantly lower revision rate in the inlay group (0.90 vs 1.61 per 100 component years; p = 0.023), suggesting tibial preparation technique may influence implant longevity (Table 4). This may reflect better cortical rim preservation with inlay preparation, promoting enhanced fixation whilst preserving bone stock. Additionally, time to revision was shorter in the onlay group (5.7 years vs 8.3 years).

While inlay preparation demonstrated superior survivorship, it is important to note that surgeon experience and compartment involvement differed between groups. Inlay procedures were more commonly primarily performed by consultants, whereas onlay procedures had greater involvement by a trainee under direct consultant supervision. Furthermore, lateral UKA was more frequent in the onlay group. Both factors – seniority of primary surgeon and compartment involved – may independently influence revision risk and should be considered as potential confounders.

Previous studies evaluating inlay and onlay techniques in isolation have shown good long-term outcomes, but few directly compare the two techniques.^14,16 To our knowledge, this is the first national registry-based study to evaluate inlay versus onlay tibial preparation technique for all-polyethylene UKA in a real-world cohort.

Limitations

Our findings must be interpreted with caution and are applicable only to fixed-bearing all-polyethylene tibial component UKA designs similar to the design evaluated in this study. The authors acknowledge the limitations of a retrospective registry study design. Inlay procedures were more often primarily performed by consultants and lateral UKA procedures, which may partially explain the observed differences in revision rates. Unmeasured confounders including BMI, radiographic progression, surgeon volume and nuances related to surgical technique were not available and may further contribute to unmeasured confounding. Nonetheless, this is the first study to compare inlay and onlay all-polyethylene tibial UKA using long-term registry data.

Conclusion

UKA with a fixed bearing all-polyethylene tibial component demonstrates favorable long-term survivorship in individuals over 75 years old. Inlay tibial preparation was associated with a signficiantly lower revision risk compared to onlay, potentially due to improved cortical rim preservation and fixation stability.

These findings support the use of all-polyethylene inlay UKA in carefully selected elderly patients with good bone quality and lower activity demands. Results should be interpreted with consideration of registry limitations and potential confounders.

Footnotes

ORCID iD

Katarina Sim

Funding

The authors received no financial support for the research, authorship, and/or publication of this article.

Declaration of conflicting interests

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

References

Kleeblad

Zuiderbaan

Hooper

, et al. Unicompartmental knee arthroplasty: state of the art. Journal of ISAKOS 2017; 2(2): 97–107.

McKie

Willis

Frampton

, et al.

NEW ZEALAND JOINT REGISTRY 25-Year report

2024. https://www.nzoa.org.nz/nzoa-joint-registry

Rodríguez-Merchán

Gómez-Cardero

. Unicompartmental knee arthroplasty: current indications, technical issues and results. EFORT Open Rev 2018; 3(6): 363–373.

Bruni

Iacono

Akkawi

, et al. Unicompartmental knee replacement: a historical overview. Joints 2013; 1(2): 45–47.

Johal

Nakano

Baxter

, et al. Unicompartmental knee arthroplasty: the past, current controversies, and future perspectives. J Knee Surg 2018; 31(10): 992–998.

Brown

Sheth

Davis

, et al. Total knee arthroplasty has higher postoperative morbidity than unicompartmental knee arthroplasty: a multicenter analysis. J Arthroplast 2012; 27(8 Suppl): 86–90.

Tripathy

Varghese

Srinivasan

, et al. Joint awareness after unicompartmental knee arthroplasty and total knee arthroplasty: a systematic review and meta-analysis of cohort studies. Knee Surg Sports Traumatol Arthrosc 2021; 29(10): 3478–3487.

Zuiderbaan

van der List

Khamaisy

, et al.

Unicompartmental knee arthroplasty versus total knee arthroplasty: which type of artificial joint do patients forget?

Knee Surg Sports Traumatol Arthrosc 2017; 25(3): 681–686.

Kim

Koh

Choi

, et al. Differences in patient-reported outcomes between unicompartmental and total knee arthroplasties: a propensity score-matched analysis. J Arthroplast 2017; 32(5): 1453–1459.

10.

Morris

Molli

Berend

, et al. Mortality and perioperative complications after unicompartmental knee arthroplasty. Knee 2013; 20(3): 218–220.

11.

Costa

Lo Presti

Grassi

, et al. Metal-backed tibial components do not reduce risk of early aseptic loosening in unicompartmental knee arthroplasty: a systematic review and meta-analysis. J Knee Surg 2020; 33(2): 180–189.

12.

Walker

Parakh

Chaudhary

, et al. Comparison of interface stresses and strains for onlay and inlay unicompartmental tibial components. J Knee Surg 2011; 24(2): 109–115.

13.

Lee

Cheng

Chen

, et al. No difference in functional outcomes, quality of life and survivorship between metal-backed and all-polyethylene tibial components in unicompartmental knee arthroplasty: a 10-year follow-up study. Knee Surg Sports Traumatol Arthrosc 2021; 29(10): 3368–3374.

14.

Bruce

Hassaballa

Robinson

, et al. Minimum 10-year outcomes of a fixed bearing all-polyethylene unicompartmental knee arthroplasty used to treat medial osteoarthritis. Knee 2020; 27(3): 1018–1027.

15.

Gomberg

Marullo

Leal-Blanquet

, et al.

European society of sports traumatology, knee surgery and arthroscopy unicompartmental knee arthroplasty manual

2023.

16.

Lustig

Paillot

Servien

, et al. Cemented all polyethylene tibial insert unicompartimental knee arthroplasty: a long term follow-up study. Orthop Traumatol Surg Res 2009; 95(1): 12–21.

17.

Rahardja

Allan

Frampton

, et al. Completeness and capture rate of publicly funded arthroplasty procedures in the New Zealand Joint Registry. ANZ J Surg 2020; 90(12): 2543–2548.

18.

Liddle

Pandit

Judge

, et al. Optimal usage of unicompartmental knee arthroplasty: a study of 41 986 cases from the national joint registry for England and wales. Bone Joint J 2015; 97-B(97): 1506–1511.

19.

Stone

Nugent

Young

, et al. The lifetime risk of revision following total knee arthroplasty: a New Zealand Joint Registry Study. Bone Joint J 2022; 104-B(2): 235–241.

20.

Tay

McGlashan

Monk

, et al. Revision indications for medial unicompartmental knee arthroplasty: a systematic review. Arch Orthop Trauma Surg 2022; 142(2): 301–314.

21.

Baker

Petheram

Avery

, et al. Revision for unexplained pain following unicompartmental and total knee replacement. J Bone Joint Surg Am 2012; 94(17): e126.

22.

Pearse

Hooper

Rothwell

, et al. Survival and functional outcome after revision of a unicompartmental to a total knee replacement: the New Zealand national joint registry. J Bone Joint Surg Br 2010; 92(4): 508–512.

23.

Tay

Young

Frampton

, et al. The lifetime revision risk of unicompartmental knee arthroplasty. Bone Joint J 2022; 104-B(6): 672–679.

24.

Levy

Fusco

Salazar-Restrepo

, et al. Unicompartmental knee arthroplasty revised to total knee arthroplasty versus primary total knee arthroplasty: a meta-analysis of matched studies. Knee 2023; 45: 1–10.

25.

Jonas

Shah

Mitra

, et al. 5-year cost/benefit analysis of revision of failed unicompartmental knee replacements (UKRs); not ‘just’ a primary total knee replacement (TKR). Knee 2014; 21(4): 840–842.

26.

Della Valle

. Revising an HTO or UKA to TKA: is it more like a primary TKA or a revision TKA? J Arthroplast 2014; 29(9): 229–231.

27.

Leta

Lygre

SHL

Skredderstuen

, et al. Outcomes of unicompartmental knee arthroplasty after aseptic revision to total knee arthroplasty a comparative study of 768 TKAs and 578 UKAs revised to TKAs from the Norwegian arthroplasty register (1994 to 2011). J Bone Joint Surg Am 2016; 98(6): 431–440.

28.

Newman

Pydisetty

Ackroyd

. Unicompartmental or total knee replacement: the 15-year results of a prospective randomised controlled trial. J Bone Joint Surg Br 2009; 91(1): 52–57.

29.

van Der List

Zuiderbaan

Pearle

. Why do medial unicompartmental knee arthroplasties fail today? J Arthroplast 2016; 31: 1016–1021.

30.

Hawi

Plutat

Kendoff

, et al. Midterm results after unicompartmental knee replacement with all-polyethylene tibial component: a single surgeon experience. Arch Orthop Trauma Surg 2016; 136(9): 1303–1307.

31.

Chen

Tay

Bolam

, et al. Accuracy and completeness of registry-reported unicompartmental knee arthroplasty revision. ANZ J Surg 2024; 94(9): 1502–1506.

32.

Bruni

Gagliardi

Akkawi

, et al. Good survivorship of all-polyethylene tibial component UKA at long-term follow-up. Knee Surg Sports Traumatol Arthrosc 2016; 24(1): 182–187.

33.

Fricka

Wilson

Strait

, et al. Outcomes of fixed versus mobile-bearing medial unicompartmental knee arthroplasty. Bone Joint J 2024; 106-B(9): 916–923.

Outcome of a Fixed-Bearing all-polyethylene unicompartmental tibial tray: A 12-year Joint Registry Follow-Up study

Abstract

Aims

Methods

Results

Conclusion

Keywords

Introduction

Methods

Results

Discussion

UKA survivorship and age

UKA revision indications

Inlay versus onlay tibial implantation

Limitations

Conclusion

Footnotes

ORCID iD

Funding

Declaration of conflicting interests

References