Comparing the mid-vastus and medial parapatellar approaches in total knee arthroplasty: A meta-analysis of short term outcomes

The Knee 19 (2012) 229–236

Contents lists available at ScienceDirect

The Knee

Review

Comparing the mid-vastus and medial parapatellar approaches in total knee arthroplasty: A meta-analysis of short term outcomes Ilhan Alcelik a, Mohamed Sukeik b,⁎, Raymond Pollock c, Anand Misra a, Ashraf Naguib a, Fares S. Haddad b a b c

Department of Trauma & Orthopaedics, West Cumberland Hospital, Whitehaven, CA28 8JG, UK Departments of Trauma and Orthopaedics, University College London Hospital, London, NW1 2BU, UK Medistats, Gledsnest Hawick, TD9 0LF, UK

a r t i c l e

i n f o

Article history: Received 24 January 2011 Received in revised form 10 July 2011 Accepted 25 July 2011 Keywords: Mid-vastus Medial parapatellar Approach Total knee arthroplasty

a b s t r a c t Purpose: Proponents of a mid-vastus (MV) approach for primary total knee arthroplasties (TKA) stress its importance in preserving function of the extensor mechanism with earlier rehabilitation and decreased prevalence of lateral release. We conducted a meta-analysis of randomised and quasi-randomised controlled trials to compare functional outcomes of the standard medial parapatellar (PP) and mid-vastus (MV) approaches in primary knee arthroplasties. Methods: The study was conducted according to the guidelines described in the Cochrane Handbook for Systematic Reviews of Interventions. Methodological features were rated independently by two reviewers. Results: We included 18 studies involving 1040 patients with mean age of 69 (SD ± 2.7) for the MV and 68.8 (SD ± 2.8) for the PP group. Using a MV approach led to significant improvement in flexion (mean difference (MD) 8.88, 95% confidence interval (CI) 4.50 to 13.25, P b 0.01) and visual analogue scale score (MD − 1.72 95% CI − 2.08 to − 1.36, P b 0.01) in the first week postoperatively and reduced the number of required lateral releases (risk difference − 0.16 95% CI − 0.30 to − 0.01, P = 0.03) with no increase in complication rates. Conclusion: We conclude that the MV approach may provide an alternative to the standard PP approach with earlier rehabilitation and decreased lateral release rates in primary TKA and no increase in complication rates. © 2011 Elsevier B.V. All rights reserved.

Contents 1. 2. 3.

4.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . Materials and methods . . . . . . . . . . . . . . . . . . . Study selection criteria . . . . . . . . . . . . . . . . . . . 3.1. Types of studies . . . . . . . . . . . . . . . . . . . 3.2. Types of participants . . . . . . . . . . . . . . . . . 3.3. Types of interventions . . . . . . . . . . . . . . . . 3.4. Types of outcome measures . . . . . . . . . . . . . 3.5. Search methods for identification of studies . . . . . . 3.5.1. Data collection and analysis . . . . . . . . . 3.6. Assessment of methodological quality of included studies 3.7. Data extraction and management . . . . . . . . . . . 3.8. Statistical analysis . . . . . . . . . . . . . . . . . . 3.8.1. Subgroup analysis . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.1. Description of studies . . . . . . . . . . . . . . . . 4.2. Effects of interventions . . . . . . . . . . . . . . . . 4.2.1. Postoperative knee flexion . . . . . . . . . . 4.2.2. Postoperative pain scores . . . . . . . . . . 4.2.3. Days to straight leg raise (SLR) . . . . . . . . 4.2.4. Functional outcome scores . . . . . . . . . . 4.2.5. Lateral releases . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . . .

230 230 230 230 230 230 230 231 231 231 232 232 232 232 232 233 233 233 233 233 233

⁎ Corresponding author at: Clinical Research Fellow in Orthopaedics, University College London Hospital, London, NW1 2BU, UK. Tel.: + 44 7530271137; fax: + 44 2079082060. E-mail address: [email protected] (M. Sukeik). 0968-0160/$ – see front matter © 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.knee.2011.07.010

230

I. Alcelik et al. / The Knee 19 (2012) 229–236

4.2.6. Duration of surgery . 4.2.7. Length of Hospital stay 4.2.8. Perioperative blood loss 4.2.9. Complications . . . . 5. Discussion . . . . . . . . . . . . . Conflict of interest and funding . . . . . . References . . . . . . . . . . . . . . . .

. . . . . . . . . . . . . . . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

1. Introduction A standard medial parapatellar approach for TKAs provides excellent exposure to the joint but also violates the extensor mechanism and blood supply to the patella [1]. The mid-vastus approach has been popularised as an alternative to the PP approach. Advocates of this technique suggest decreased post-operative pain, earlier functional recovery, shorter hospital stay and reduced blood loss and patellar complications relating to devascularisation of the patella such as patellar fracture, button loosening and anterior knee pain [2]. Potential disadvantages however, may include reduced overall exposure and damage to neurovascular structures. Nowadays, around 3% of primary cemented TKAs in England and Wales are performed using a MV approach, whereas in 93% of cases, a standard PP approach is still preferred [3]. In the current study, we hypothesised that there are significant benefits in using the MV as opposed to the standard PP approach in the early rehabilitation period after primary TKAs. The hypothesis was tested using a meta-analysis of randomised and quasi-randomised controlled trials comparing functional outcomes of the study groups after primary TKA. 2. Materials and methods A systematic review and meta-analysis was conducted according to the guidelines described in the Cochrane Handbook for Systematic Reviews of Interventions and PRISMA Statement [4,5]. 3. Study selection criteria 3.1. Types of studies Randomised and quasi-randomised (e.g. allocation by hospital number or date of birth) controlled trials were included in this study. 3.2. Types of participants The participants were adult patients who underwent primary TKA using either a MV or PP approach regardless of the type or size of prosthesis. 3.3. Types of interventions The intervention considered was using a MV versus standard PP approach. We have not excluded studies where a minimally invasive approach was utilised (without concomitant use of computer navigation) but instead performed subgroup analysis where appropriate to detect any relevant effect on outcomes. 3.4. Types of outcome measures The primary outcome measure was postoperative knee flexion. The secondary outcome measures were: 1. postoperative pain scores 2. days to straight leg raise (SLR)

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

3. 4. 5. 6. 7. 8.

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

. . . . . . .

233 233 234 234 235 236 236

functional knee and quality of life outcome measures number of lateral releases duration of surgery duration of hospital stay perioperative blood loss complications such as thromboembolic events and other local and systemic complications.

Exclusion criteria were: studies not randomised, any study comparing the two approaches in dissimilar circumstances (e.g.

Table 1 Quality assessment items and possible scores. A. Was the assigned treatment adequately concealed prior to allocation? 2 = method did not allow disclosure of assignment 1 = small but possible chance of disclosure of assignment or unclear 0 = quasi-randomised or open list/tables B. Were the outcomes of participants who withdrew described and included in the analysis (intention to treat)? 2 = withdrawals well described and accounted for in analysis 1 = withdrawals described and analysis not possible 0 = no mention, inadequate mention, or obvious differences and no adjustment C. Were the outcome assessors blinded to treatment status? 2 = effective action taken to blind assessors 1 = small or moderate chance of unblinding of assessors 0 = not mentioned or not possible D. Were the treatment and control group comparable at entry? (Likely confounders may be age, partial or total rupture, activity level, acute or chronic injury.) 2 = good comparability of groups, or confounding adjusted for in analysis 1 = confounding small; mentioned but not adjusted for 0 = large potential for confounding, or not discussed E. Were the participants blind to assignment status after allocation? 2 = effective action taken to blind participants 1 = small or moderate chance of unblinding of participants 0 = not possible, or not mentioned (unless double-blind), or possible but not done F. Were the treatment providers blind to assignment status? 2 = effective action taken to blind treatment providers 1 = small or moderate chance of un blinding of treatment providers 0 = not possible, or not mentioned (unless double-blind), or possible but not done G. Were care programmes, other than the trial options, identical? 2 = care programmes clearly identical 1 = clear but trivial differences 0 = not mentioned or clear and important differences in care programmes H. Were the inclusion and exclusion criteria clearly defined? 2 = clearly defined 1 = inadequately defined 0 = not defined I. Were the interventions clearly defined? 2 = clearly defined interventions are applied with a standardised protocol 1 = clearly defined interventions are applied but the application protocol is not standardised 0 = intervention and/or application protocol are poorly or not defined J. Were the outcome measures used clearly defined? (by outcome) 2 = clearly defined 1 = inadequately defined 0 = not defined K. Were diagnostic tests used in outcome assessment clinically useful? (by outcome) 2 = optimal 1 = adequate 0 = not defined, not adequate L. Was the surveillance active, and of clinically appropriate duration? 2 = active surveillance and appropriate duration 1 = active surveillance, but inadequate duration 0 = surveillance not active or not defined

I. Alcelik et al. / The Knee 19 (2012) 229–236

different anaesthesia for each group of patients), animal studies and studies where the above mentioned outcomes were not evaluated. Studies where navigation assisted surgery was performed were also excluded to decrease heterogeneity amongst the studies especially when a minimally invasive approach was utilised concomitantly. 3.5. Search methods for identification of studies

Identification

The following databases were searched in February 2011 to establish whether there has been any previous systematic reviews or meta-analyses comparing MV and PP approaches in TKA: Cochrane Database of Systematic Reviews (CDSR), Database of Abstracts of Reviews of Effects (DARE), Health Technology Assessment Database (HTA), NHS Economic Evaluation Database (NHS EED), Turning Research Into Practice (TRIP) Database, Health Services/Technology Assessment Text (HSTAT), Aggressive Research Intelligence Facility (ARIF) appraisals, Scottish Intercollegiate Guideline Network Guidelines (SIGN), National Research Register (NRR), and MEDLINE (1950 to February 2011). The following exploded MeSH terms were used for the literature search: “Midvastus”, “Mid-vastus”, “Medial parapatellar”, “Median parapatellar”, “Minimally invasive” and “Total Knee”. Text searches of key fields were included. A MEDLINE search was then refined to clinical trials and randomised controlled trials (RCTs) in human adults. The search was extended to other data bases, namely EMBASE, the Cochrane Controlled Trials Register, AMED and CINAHL, Google and Google Scholar for trials of midvastus approach and total knee replacement published in any language from 1966 to February 2011. The bibliographies of retrieved trials and other relevant publications, including reviews and meta-analyses, were examined for additional articles. The following websites were searched to identify unpub-

Records identified through database searching (n = 456)

231

lished and ongoing studies: Current Controlled Trials (www. controlled-trials.com); Centre Watch (www.centerwatch.com); Trials Central (www.trialscentral.org); OpenSIGLE (System for Information on Grey Literature in Europe); UKCRN Portfolio Database; The UK National Research Register (www.nihr.ac.uk/Pages/NRRArchive. aspx). Journal of Bone and Joint Surgery — British Volume and American Volume (www.ejbjs.org), and the American Academy of Orthopaedic Surgeons (www.aaos.org) were searched manually.

3.5.1. Data collection and analysis 3.5.1.1. Selection of the studies. Three authors (IA, MS and RP) applied the search strategy independently and all relevant study abstracts were hand searched by them after which potentially suitable studies were reviewed in full paper format by each of the authors independently. Authors were contacted for more information and clarification of data as necessary. Disagreement was discussed with the senior authors and when no consensus was reached, the particular study was excluded.

3.6. Assessment of methodological quality of included studies The review authors used a modification of the generic evaluation tool used by the Cochrane Bone, Joint and Muscle Trauma Group [6] (Table 1). Two authors (RP and AM) assessed the methodological quality of each study. Disagreement was resolved by the senior authors. Although the total quality assessment score (QAS) was reported for each study, it was not used to weight the studies in the meta-analysis.

Additional records identified through other sources (n = 96)

Eligibility

Screening

Records after duplicates removed (n = 481)

Records screened (n = 481)

Records excluded (n = 457)

Full-text articles assessed for eligibility (n = 24)

Full-text articles excluded, with reasons (n = 6)

Included

Studies included in qualitative synthesis (n = 18)

Studies included in quantitative synthesis (meta-analysis) (n = 18)

Fig. 1. PRISMA chart of the study selection process.

232

I. Alcelik et al. / The Knee 19 (2012) 229–236

3.7. Data extraction and management A data extraction form was designed and agreed by the authors. A pilot test of five articles was performed to ensure the form's consistency. Initially, two authors (IA and MS) extracted the data independently which was later on reviewed jointly to produce agreed accurate data. Disagreements were resolved by consensus or consultation with the senior authors. Authors of individual trials were contacted directly to provide further information when necessary.

heterogeneity, prompting random effects modelling estimate; otherwise a fixed effects approach was used. On the other hand, a nonsignificant chi-squared test result only suggested that there is no evidence of heterogeneity. It did not imply that there was necessarily homogeneity as there may have been insufficient power to be able to detect heterogeneity. 3.8.1. Subgroup analysis When data allowed, we performed sub-group analysis of the trials according to the type of prosthesis used and whether a minimally invasive or standard incision was implemented.

3.8. Statistical analysis 4. Results

Meta-analysis, performed by Review Manager (version 5.0, Nordic Cochrane Center, Copenhagen), was used to combine the relevant estimates of the effect of interest from the selected studies to provide an overall estimate of the effect. In five studies [7–11], missing standard deviations were calculated using the range of values given as suggested by Hozo et al. [12]. Continuous data for each arm in a particular study was expressed as mean and standard deviation and the treatment effect as mean differences. Dichotomous data for each arm in a particular study was expressed as proportions or risks, and the treatment effect as risk differences. For dichotomous data, the Mantel–Haenszel method was used to combine the estimates, whereas for continuous data the inverse variance method was utilised. Missing data was sought from the authors. Where this was not possible or data was missing through loss to follow-up, intentionto-treat principles were used. Statistical heterogeneity was assessed using the result of the chisquared test. A p value of b0.1 was considered suggestive of statistical

4.1. Description of studies Four hundred and eighty one studies were identified. Four hundred fifty seven were excluded based on the inclusion/exclusion criteria, leaving 24 potentially relevant papers for detailed evaluation (Fig. 1). This was further reduced to 18 studies after critical appraisal of the full papers; Table 2 shows the included studies and data extracted [1,2,7–11,13–23]. Reasons for further exclusion included irrelevant outcomes measured in individual studies such as postoperative symptomatic transient patellar ischaemia [24], lateral retinacular tension [25] and alteration of patellar tracking by measuring patellar tilt [26]. Parentis et al. [27] published a longer term follow-up of the same patients (Kelly et al.) [1] and therefore, their earlier study was also excluded. The remaining two studies excluded were neither RCTs nor quasirandomised trials [28,29]. Two studies were quasi-randomised trials [1,2]. The majority of studies were small with participant numbers ranging from 39 to 106. However, they were relatively well designed with a mean QAS of 18 out of 24 (range 11–22). Only two studies had a score of less than 15 [1,21]. A funnel plot based on the most frequently cited outcome was broadly symmetrical indicating minimal publication bias (Fig. 2). Participants of all studies were adult patients who underwent primary TKA. The main indications for an operation were osteoarthritis and rheumatoid arthritis.

Table 2 Data extracted. Standard deviations are given between the brackets and an asterisk (*) signifies indirect measurement of it from the range. h = hours, d = days, MV = mid-vastus, PP = medial para-patellar, SLR = straight leg raise, blood loss: T = total, D = drain, I = intraoperative, E = estimated.

Bathis 2005 MV PP Chin 2007 MV PP Dalury 2008 MV PP Engh 1997 MV PP Hernandez MV Vaquero 2010 PP Jousponis 2008 MV PP Jung 2007 MV PP Kang 2004 MV PP Karachalios 2008 MV PP Karpman 2009 MV PP Keating 1999 MV PP Kelly 2006 MV PP Kim 2010 MV PP Kolisek 2007 MV PP Nestor 2010 MV PP Pei-liang 2008 MV PP Schneider 1998 MV PP Walter MV PP

Duration of surgery

Days to SLR

Lateral release

Measured blood loss

Hospital stay

Thromboembolism

Other complications

80.6 (17.4) 79.1 (16.4) 111.7 (21.2)* 92 (16.2)* – – – – 130 (26.94) 103.19 (20.94) 93 (8.7) 86 (6.9) 87.1 (13.4) 83.5 (15) – – 75 (7.5)* 55 (6.25)* 67 (8.1) 65 (6.4) – – – – – – – – – – 61.4 (11.2) 60.5 (10.6) – – – –

– – – – – – 3.4 (2.1) 3.9 (2.3) – – – – 1.8 (4.64) 2.1 (5.03) – – – – – – Not suitable – – – – – – – – 1.8 (0.3) 4.5 (0.8) 4 (2.9) 7.1 (3.3) 1.53 (0.7) 1.38 (0.58)

– – – – – – 2 4 – – – – – – 5 22 – – – – 25 26 1 13 – – – – – – 4 7 – – 0 0

– – 411 (317)*T 403 (237)*T – – – – 621 (412)D 1072 (440)D – – 981.5 (50.2)T 1020 (51.6)T – – 613 (330)*D 1016 (250)*D 648 (385)E 623 (241)E – – Not suitable – – 108 (93.7)*I 110 (70)*I – – 286 (14.9)*E 368 (17.9)*E – – – –

– – 6 d (2.7)* 6.1 d (2)* – – – – 6.92 d (1.47) 7.88 d (2.06) – – – – – – Not suitable 70 h (16.7) 68 h (14.9) – – – – – – Not suitable – – – – Not suitable 2.58 d (1.17) 2.48 d (0.89)

– – 0 2 – – 2 2 – – – – – – – – 1 2 – – – – 1 0 – – 0 1 2 4 – – 1 2 0 0

– – 1 0 – – – – 6 2 – – 2 0 – – 6 4 – – 3 0 – – – – 7 2 3 7 0 1 1 5 0 0

I. Alcelik et al. / The Knee 19 (2012) 229–236

233

but with significant statistical heterogeneity (P = 0.07). Results for group two were statistically insignificant (P = 0.11) (Fig. 3). There have also been individualised attempts to compare flexion amongst the groups at longer periods of follow-up. For example, Pei-Liang [11] assessed flexion at 90 days (MV 109 SD ± 18 vs PP 106 SD ±20) and Karachalios [9] at 23 months (MV 126.5 SD± 10 vs PP 116 SD±10). However, there has not been enough data to perform a metaanalysis of medium to long term results.

Fig. 2. Funnel plot demonstrating minimal publication bias from complications outcome.

However, a few patients had sero-negative arthritis as an underlying pathology. MV approach was used in 188 male and 336 female patients whilst 178 male and 338 female patients underwent a PP approach. The mean age of patients was 69 (SD ± 2.7) for the MV and 68.8 (SD ± 2.8) for the PP group. Management protocols were similar and patients' demographics were comparable for each study as detailed in Table 3. A cruciate retaining prosthesis was used in three studies [13–15], posterior stabilised prosthesis in 8 studies [1,7,10,18–22] whereas the rest did not mention the type of prosthesis used. Nine of the studies used a minimally invasive surgical approach for the MV group of patients [7,9,10,13,15,18–20,22] and the rest used a standard midline incision. Tourniquet application was mentioned in 6 studies [1,10,11,14,15,19]. There were 3 studies which mentioned using solely regional anaesthesia [14,19,22]. A drain was routinely used in 8 studies [7,9,14,15,19–22]. Deep venous thrombosis (DVT) prophylaxis regime was mentioned in 4 studies only and included low molecular weight heparin (LMWH) [9,15], warfarin and aspirin [13], or a combination of warfarin/aspirin or warfarin/LMWH [20]. DVT screening protocol was described in two studies and included routine ultrasound in the postoperative period [2,9].

4.2. Effects of interventions 4.2.1. Postoperative knee flexion Seven studies [2,8,11,13,14,19,22] were suitable for the meta-analysis of postoperative knee flexion in 222 MV versus 217 PP knees. Knee flexion was assessed at different intervals after the operation in each study. Accordingly, we compared the results for measurements available within one week (group 1) and 6 weeks postoperatively (group 2). If more than one measurement were performed in the first week after the operation, the latest was considered for the analysis. Two other studies [1,23] included the mean values of flexion only and therefore, it was impossible to analyse their data. Post-operative knee flexion was significantly greater in the MV compared to the PP approach in group 1 with MD of 8.88 (95% CI 4.50 to 13.25, P b 0.01)

4.2.2. Postoperative pain scores Three studies [11,13,14] were suitable for assessment of postoperative pain using the visual analogue scale (VAS) scores in 79 MV and 78 PP knees. Of note is that the type of anaesthesia and postoperative pain management protocols were identical for each group within a study. Pain was also assessed at different intervals after the operation in each study. Accordingly, we compared the results for measurements available at days 1 (group 1), 3 (group 2) and 15 (group 3) postoperatively. Two other studies [9,28] included the mean values of VAS only and therefore; it was impossible to include their data. Post-operative VAS was significantly lower in the MV compared to the PP approach in all groups especially group 2 with MD −0.67 (95% CI − 0.98 to − 0.36, P b 0.01) in group 1, MD − 1.72 (95% CI − 2.08 to − 1.36, P b 0.01) in group 2 and MD − 0.18 (95% CI − 0.32 to − 0.05, P b 0.01) for group 3 with no evidence of statistical heterogeneity (P = 0.17, 0.26 and 0.40 respectively) (Fig. 4). 4.2.3. Days to straight leg raise (SLR) Five studies were eligible for this outcome including 201 MV and 185 PP patients [2,11,17,20,21]. Results for this outcome were statistically insignificant (P = 0.14). 4.2.4. Functional outcome scores Three studies [8,10,22] reported the functional Knee Society Score (KSS) in 380 patients at 6 and 12 weeks postoperatively. Again results did not reach statistical significance (P = 0.24 and P = 0.69 respectively). 4.2.5. Lateral releases The number of lateral releases was reported in six studies [1,2,11,16,20,23] which included 291 MV and 271 PP cases. The results showed a significant reduction in the number of lateral releases in the MV group (37 vs. 72) with risk difference (RD) −0.16 (95% CI − 0.30 to − 0.01, P = 0.03) but also significant statistical heterogeneity (P b 0.01) (Fig. 5). 4.2.6. Duration of surgery Eight studies [7,9,11,13–15,17,22] were eligible for this outcome reporting duration of surgery on 269 MV and 278 PP cases. Four studies [1,10,28,29] mentioned the tourniquet time instead. However; those were not included as this was clearly different to the overall duration of surgery. Duration of surgery was significantly shorter in the PP group with MD 9.63 min (95% CI 2.78 to 16.48, P b 0.01). However, there has been significant statistical heterogeneity amongst the studies included (P b 0.01) (Fig. 6). 4.2.7. Length of Hospital stay Four studies reported length of hospital stay [7,13,15,20]. Three further studies [9,10,21] quoted the means only and therefore were not included in the analysis. Results showed no significant difference amongst the study groups in relation to this outcome (P = 0.80).

Table 3 Characteristics of the included studies, MV: mid-vastus, PP: medial para-patellar, n: number of patients, M: male, F: female, BMI: body mass index, N/S: not specified. MV

Bathis Chin Dalury Engh Hernandez-Vaquero Jousponis Jung Kang Karachalios Karpman Keating Kelly Kim Kolisek Nestor Pei-liang Schneider Walter Total/mean

PP

n

Knees

M

F

Age

BMI

n

Knees

M

F

Age

BMI

25 30 20 61 26 35 49 32 50 20 100 20 23 40 27 34 14 42 648

25 30 20 61 26 35 49 32 50 20 100 22 23 40 27 34 15 42 651

6 6 6 25 5 5 N/S N/S 19 7 40 N/S N/S 29 9 7 14 10 188

19 24 14 36 21 30 N/S N/S 31 13 60 N/S N/S 11 18 27 0 32 336

68.7 67.4 67 68 70.8 72 N/S N/S 71.1 74 70.2 68.2 67 67 N/S 70.5 71 63 69

30 28.5 30 N/S 32.1 28 N/S N/S 32 30 32.9 N/S 27.1 32 29.6 N/S N/S 35.5 30.64

25 30 20 57 36 35 49 32 50 19 100 27 22 40 27 34 25 19 647

25 30 20 57 36 35 49 32 50 19 100 29 22 40 27 34 26 19 650

6 3 6 16 6 5 N/S N/S 15 9 40 N/S N/S 24 9 7 25 7 178

19 27 14 41 30 30 N/S N/S 35 10 60 N/S N/S 15 18 27 0 12 338

70.5 63.4 67 70 70.5 71.4 N/S N/S 70.8 73 70.2 65.5 68 70 N/S 70.5 64.5 66.6 68.79

30 29.4 30 N/S 30.8 29 N/S N/S 31.5 29 32.9 N/S 28.4 30 29.6 N/S N/S 34.8 30.45

234

I. Alcelik et al. / The Knee 19 (2012) 229–236

Fig. 3. Flexion outcome forest plot analyses at 1 and 6 weeks. The black diamond signifies that the mean difference is in favour of the midvastus approach. The size of each square depends on the weight of each study as detailed in the forest plot. A green coloured square is given to continuous outcomes and a blue square to dichotomous outcomes. MV = midvastus approach, PP = medial parapatellar approach. 4.2.8. Perioperative blood loss This outcome has not been meta-analysed as strategies for measuring blood loss were different in each study such as reporting intra-operative blood loss, total blood loss, blood in the drains and estimated blood loss and therefore, the groups were not like for like to allow comparison [7,9–11,13,15,17]. 4.2.9. Complications 1. Thromboembolic events: twelve trials [1,2,7–11,13,19,21–23] reported on DVT and pulmonary embolism (PE) with a total number of 921 patients of whom 454 underwent a MV approach. There were overall six reported cases of DVT and one PE in the MV group, 11 cases of DVT and two cases of PE in the PP group. However, the results were statistically insignificant. (RD − 0.01, 95% CI − 0.03 to 0.01, P = 0.31). 2. Other complications: in this section, we compared all other reported complications such as haematoma formation, secretion from drain sites, postoperative stiffness and systemic complications. Systemic complications reported included one case of myocardial infarction which recovered uneventfully and a patient who had

temporary mental changes postoperatively in the MV group. In the PP group, one patient had congestive heart failure but made a full recovery and two patients had pneumonia treated with no further complications. Overall results showed that a PP approach was associated with fewer such complications. However, this has not reached a statistically significant level (RD 0.02 95% CI − 0.01 to 0.04, P = 0.13). Two studies [1,8] compared the electromyographic findings of the quadriceps muscle postoperatively in the study groups. Whilst no abnormalities were found in the PP group, Kelly et al. [1] reported nine of 21 knees having abnormal findings in the MV group at 6 months postoperatively. However, in seven cases where the vastus split had been developed bluntly, the findings normalised at 5 years follow-up. The other two knees with chronic changes at 5 years had sharp dissection for the muscle split, but neither groups demonstrated any functional compromise. The authors concluded that such findings therefore, most likely represented reversible neuropraxic injury of no functional sequelae that may be avoided by blunt dissection in the vastus medialis muscle. Dalury et al. [8] reported abnormal electromyographic findings in two of 20 patients in the MV group and one of 20

Fig. 4. Visual analogue scale (VAS) outcome forest plot analyses at 1, 3 and 15 days.

I. Alcelik et al. / The Knee 19 (2012) 229–236

235

Fig. 5. Number of lateral releases and forest plot analysis. patients in the PP group at 6 weeks postoperatively which resolved spontaneously at 12 weeks follow-up. Subgroup analysis according to prosthesis type and whether a minimally invasive approach has been utilised showed similar trends for all outcomes measured.

5. Discussion Since its inception, PP approach has been the most popular approach used for TKAs due to the combination of simplicity, familiarity and excellent exposure [3,30]. However, critics of this approach argue that incising the quadriceps tendon weakens quadriceps function and adversely affects the blood supply to the patella. More recently, the MV approach gained popularity due to preservation of the quadriceps tendon and possible earlier return of the quadriceps function. Despite the longer duration of surgery, our most significant findings showed that a MV approach improved postoperative pain and knee flexion in the first week after a knee replacement. White et al. [29] suggested that such earlier recovery may be associated with shorter hospital stay which would also have an impact on the overall cost of the TKA. However, our findings did not support such a relationship. Despite trials consistently suggesting benefit, there was significant heterogeneity between findings for most of the outcomes measured. Variations which may have accounted for such heterogeneity include the following: 1. the difference in sample sizes 2. the variation of patients' demographics such as age, gender, BMI and severity of the underlying illness (Table 2) 3. inclusion and exclusion criteria for each study 4. the differences in management protocols between treating centres including the surgical technique such as patellar eversion, usage of drains or tourniquets, type of anaesthesia, DVT prophylaxis and postoperative analgesic regimes and rehabilitation programmes. Engh et al. [31] reported on 176 patients who underwent TKA using a PP or MV approach and concluded that 50% of the PP group needed lateral releases whereas only 3% did in the MV group. They

related their findings to the quadriceps tendon disruption which resulted in a high number of lateral releases in the PP group. Our study showed similar trends but less pronounced results with 26.5% PP vs 12.7% MV patients requiring a lateral release. The only systematic review we found in the literature examining the relationship between the MV and PP approaches was published in 2010 by Ying-long et al. [32]. However, their search yielded eight studies only; four RCTs, two quasi-randomised studies and two non RCTs which affected the strength of point estimates and conclusions drawn from the study. There are several strengths of this meta-analysis. First, the subject has never been investigated to this depth before and there are no relevant Cochrane reviews in the literature. Second, we conducted a thorough literature search of randomised controlled trials, including publications in any language as well as unpublished abstracts. Third, the results of the study showed significant improvement of knee flexion and VAS scores in the first week postoperatively for the MV group which supports our hypothesis that a MV approach is associated with earlier rehabilitation after primary TKAs with no increased risk of complication rates. Limitations of the study included insufficient data to support the analysis of medium or long term outcomes. Furthermore, there has been significant heterogeneity amongst the study groups for most outcomes measured which reflects not only the differences in surgical technique and perioperative care but also the demographics of the study population. Finally, it is also important to emphasise that there are limitations associated with using a MV approach. According to Engh who first described the operative technique; weight, hypertrophic arthritis and previous high tibial osteotomies are considered to be relative contraindications due to the limitations of exposure in comparison to the PP or midline arthrotomy. Additionally, patients with less than 80° of flexion are better operated on using a PP approach if a quadriceps snip is contemplated [33]. In summary, we conclude that the MV approach offers superior range of movement and pain control over the standard PP approach in the immediate postoperative period after TKA. Additionally, it reduces the need for lateral release with no increased risk of complications or

Fig. 6. Forest plot analysis of duration of surgery.

236

I. Alcelik et al. / The Knee 19 (2012) 229–236

impairment of quadriceps function. However, future randomised trials of sufficient power should be designed to compare the medium and long term outcomes between a MV and PP approach. Conflict of interest and funding The authors declare that they have no conflict of interest and no source of external funding was applied for in relation to this study. References [1] Kelly MJ, Rumi MN, Kothari M, Parentis MA, Bailey KJ, Parrish WM, et al. Comparison of the vastus-splitting and median parapatellar approaches for primary total knee arthroplasty: a prospective, randomized study. J Bone Joint Surg Am 2006 Apr;88(4):715–20. [2] Engh GA, Holt BT, Parks NL. A midvastus muscle-splitting approach for total knee arthroplasty. J Arthroplasty 1997 Apr;12(3):322–31. [3] The National Joint Registry for England and Wales 6th Annual Report; 2009. p. 72. [4] Higgins JPT GS. Cochrane Handbook for Systematic Reviews of Interventions Version 5.0.2. 2006 Sept [updated September 2009]. [5] Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA Statement. PLoS Med 2009 Jul 21;6(7): e1000097. [6] Handoll H, Elstub L, J E. Cochrane Bone, Joint and Muscle Trauma Group. About the Cochrane collaboration (Cochrane Review Groups (CRGS)); 2008. 4. Art. No: MUSKINJ. [7] Chin PL, Foo LS, Yang KY, Yeo SJ, Lo NN. Randomized controlled trial comparing the radiologic outcomes of conventional and minimally invasive techniques for total knee arthroplasty. J Arthroplasty 2007 Sep;22(6):800–6. [8] Dalury DF, Snow RG, Adams MJ. Electromyographic evaluation of the midvastus approach. J Arthroplasty 2008 Jan;23(1):136–40. [9] Karachalios T, Giotikas D, Roidis N, Poultsides L, Bargiotas K, Malizos KN. Total knee replacement performed with either a mini-midvastus or a standard approach: a prospective randomised clinical and radiological trial. J Bone Joint Surg Br 2008 May;90(5):584–91. [10] Kolisek FR, Bonutti PM, Hozack WJ, Purtill J, Sharkey PF, Zelicof SB, et al. Clinical experience using a minimally invasive surgical approach for total knee arthroplasty: early results of a prospective randomized study compared to a standard approach. J Arthroplasty 2007 Jan;22(1):8–13. [11] Pei-liang FX-h L, Yu-li W. Comparison of midvastus and standard medial parapatellar approaches in total knee arthroplasty. Zhongguo Zuzhi Gongcheng Yanjiu Yu Linchuang Kangfu 2008;12(9):1793–6. [12] Hozo SP, Djulbegovic B, Hozo I. Estimating the mean and variance from the median, range, and the size of a sample. BMC Med Res Methodol 2005;5(1):13. [13] Karpman RR, Smith HL. Comparison of the early results of minimally invasive vs standard approaches to total knee arthroplasty: a prospective, randomized study. J Arthroplasty 2009 Aug;24(5):681–8. [14] Bathis H, Perlick L, Blum C, Luring C, Perlick C, Grifka J. Midvastus approach in total knee arthroplasty: a randomized, double-blinded study on early rehabilitation. Knee Surg Sports Traumatol Arthrosc 2005 Oct;13(7):545–50.

[15] Hernandez-Vaquero D, Noriega-Fernandez A, Suarez-Vazquez A. Total knee arthroplasties performed with a mini-incision or a standard incision. Similar results at six months follow-up BMC Musculoskelet Disord 2010 Feb;11:27. [16] Kang SB, Yoon KS, Lee JH, HC J. The comparison of the midvastus and paramedian approach for total knee arthroplasty. J Korean Knee Soc 2004;16:39–43. [17] Jung Won-Hwa, Ha Yong-Chan, Cha M-S. A comparison of the midvastus and median parapatellar surgical approaches in total knee arthroplasty. J Korean Orthop Assoc 2007;42:354–9. [18] Kim JG, Lee SW, Ha JK, Choi HJ, Yang SJ, Lee MY. The effectiveness of minimally invasive total knee arthroplasty to preserve quadriceps strength: a randomized controlled trial. Knee. 2010 Sept; 10. [19] Nestor BJ, Toulson CE, Backus SI, Lyman SL, Foote KL, Windsor RE. Mini-midvastus vs standard medial parapatellar approach: a prospective, randomized, doubleblinded study in patients undergoing bilateral total knee arthroplasty. J Arthroplasty. 2010 Sept;25(6 Suppl):5–11, e1. [20] Walter F, Haynes MB, Markel DC. A randomized prospective study evaluating the effect of patellar eversion on the early functional outcomes in primary total knee arthroplasty. J Arthroplasty 2007 Jun;22(4):509–14. [21] Schneider JM BJ, Glockner F, Melnikoff J, Fletcher F. The vastus medialis musclesplitting approach: a technique for total knee arthroplasty. Knee 1998;5(3):183–6. [22] Juosponis R, Tarasevicius S, Smailys A, Kalesinskas RJ. Functional and radiological outcome after total knee replacement performed with mini-midvastus or conventional arthrotomy: controlled randomised trial. Int Orthop 2009 Oct;33(5):1233–7. [23] Keating EM, Faris PM, Meding JB, Ritter MA. Comparison of the midvastus musclesplitting approach with the median parapatellar approach in total knee arthroplasty. J Arthroplasty 1999 Jan;14(1):29–32. [24] Gelfer Y, Pinkas L, Horne T, Halperin N, Alk D, Robinson D. Symptomatic transient patellar ischemia following total knee replacement as detected by scintigraphy. A prospective, randomized, double-blind study comparing the mid-vastus to the medial para-patellar approach. Knee Dec 2003;10(4):341–5. [25] Komatsu T, Ishibashi Y, Otsuka H, Nagao A, Toh S. The effect of surgical approaches and tourniquet application on patellofemoral tracking in total knee arthroplasty. J Arthroplasty 2003 Apr;18(3):308–12. [26] Ozkoc G, Hersekli MA, Akpinar S, Ozalay M, Uysal M, Cesur N, et al. Time dependent changes in patellar tracking with medial parapatellar and midvastus approaches. Knee Surg Sports Traumatol Arthrosc 2005 Nov;13(8):654–7. [27] Parentis MA, Rumi MN, Deol GS, Kothari M, Parrish WM, Pellegrini Jr VD. A comparison of the vastus splitting and median parapatellar approaches in total knee arthroplasty. Clin Orthop Relat Res 1999 Oct;367:107–16. [28] Dalury DF, Jiranek WA. A comparison of the midvastus and paramedian approaches for total knee arthroplasty. J Arthroplasty 1999 Jan;14(1):33–7. [29] White Jr RE, Allman JK, Trauger JA, Dales BH. Clinical comparison of the midvastus and medial parapatellar surgical approaches. Clin Orthop Relat Res 1999 Oct;367:117–22. [30] Scuderi GR, Tenholder M, Capeci C. Surgical approaches in mini-incision total knee arthroplasty. Clin Orthop Relat Res 2004 Nov;428:61–7. [31] Engh GA, Parks NL, Ammeen DJ. Influence of surgical approach on lateral retinacular releases in total knee arthroplasty. Clin Orthop Relat Res 1996 Oct;331:56–63. [32] Xu Yinglong, Zhao Jinmin, Li Shuzhen, Ding Xiaofei, Taixiang W. Midvastus and medial parapatellar approaches in total knee arthroplasty: a systematic review. Chin J Evid-Based Med 2010;10(5):585–91. [33] Engh GA, Parks NL. Surgical technique of the midvastus arthrotomy. Clin Orthop Relat Res 1998 Jun;351:270–4.