Mitral Valve Clip for Treatment of Mitral Regurgitation: An Evidence-Based Analysis

Mohammed T Ansari; Nadera Ahmadzai; Kathryn Coyle; Doug Coyle; David Moher

. 2015 May 1;15(12):1–104.

Mitral Valve Clip for Treatment of Mitral Regurgitation: An Evidence-Based Analysis

Mohammed T Ansari ¹, Nadera Ahmadzai ¹, Kathryn Coyle ², Doug Coyle ³, David Moher ¹

PMCID: PMC4561766 PMID: 26379810

Abstract

Background

Many of the 500,000 North American patients with chronic mitral regurgitation may be poor candidates for mitral valve surgery.

Objective

The objective of this study was to investigate the comparative effectiveness, harms, and cost-effectiveness of percutaneous mitral valve repair using mitral valve clips in candidates at prohibitive risk for surgery.

Data Sources

We searched articles in MEDLINE, Embase, and the Cochrane Library published from 1994 to February 2014 for evidence of effectiveness and harms; for economic literature we also searched NHS EED and Tufts CEA registry. Grey literature was also searched.

Review Methods

Primary studies were sought from existing systematic reviews that had employed reliable search and screening methods. Newer studies were sought by searching the period subsequent to the last search date of the review. Two reviewers screened records and assessed study validity. We used the Cochrane risk of bias tool for randomized, generic assessment for non-randomized studies, and the Phillips checklist for economic studies.

Results

Ten studies including 1 randomized trial were included. The majority of the direct comparative evidence compared the mitral valve clip repair with surgery in patients not particularly at prohibitive surgical risk. Irrespective of degenerative or functional chronic mitral regurgitation etiology, evidence of effectiveness and harms is inconclusive and of very low quality. Very-low-quality evidence indicates that percutaneous mitral valve clip repair may provide a survival advantage, at least during the first 1 to 2 years, particularly in medically managed chronic functional mitral regurgitation. Because of limitations in the design of studies, the cost-effectiveness of mitral valve clips in patients at prohibitive risk for surgery also could not be established.

Limitations

Because of serious concerns of risk of bias, indirectness, and imprecision, evidence is of very low quality.

Conclusions

No meaningful conclusions can be drawn about the comparative effectiveness, harms, and cost-effectiveness of mitral valve clips in the population with chronic mitral regurgitation who are at prohibitive risk for surgery.

PLAIN LANGUAGE SUMMARY

Chronic mitral regurgitation is a long-standing abnormal leakage of blood back into 1 of the heart chambers as the heart attempts to pump blood to the rest of the body. It occurs because the mitral valve is either anatomically distorted or functionally abnormal. Severe chronic mitral regurgitation is debilitating and requires corrective cardiac surgery. But many of the 500,000 patients in North America who have chronic mitral regurgitation are elderly and for various reasons are at prohibitive risk for surgery. For such patients, an approach has been proposed that involves a catheter puncturing the skin of the groin and travelling all the way to the affected valve to deploy a device that clips and repairs the valve leaflets (a mitral valve clip). This investigation sought to compare the effectiveness and cost-effectiveness of mitral valve clips with current standards of care in patients at high or prohibitive risk for surgery. To address this uncertainty, we searched, critically appraised, and collated existing research evidence.

We found sparse, very-low-quality evidence related to mitral valve clips. Most of the evidence was in populations not particularly at high risk for surgery. Important limitations in the design, conduct, and size of the studies precluded definitive conclusions about the relative effectiveness, harms, and cost-effectiveness of mitral valve clips when compared with conservative management in patients at prohibitive risk for surgery.

BACKGROUND

Objective of Analysis

This evidence-based analysis aims to collate extant evidence regarding mitral valve clips with a view to minimizing bias to answer the following research questions:

For symptomatic patients at high or prohibitive surgical risk, what are the comparative effectiveness, harms, and cost-effectiveness of percutaneous mitral valve repair using mitral valve clips?
What are the important population effect modifiers?

Clinical Need and Target Population

Description of Condition

Mitral regurgitation (MR) is an abnormal backflow of blood into the left atrium during cardiac systole that is normally prevented by the bicuspid mitral valve. It is caused by a mitral apparatus dysfunction with potential for hemodynamic compromise. (1) The mitral apparatus is a 3-dimensional functional valvular unit that comprises the left atrial wall, mitral annulus, 2 mitral valve anteromedial and posterolateral leaflets, subvalvular chordae tendineae, papillary muscles, and adjacent left ventricular wall. (1)

Mitral regurgitation can be acute or chronic. Acute MR is a medical and surgical emergency and is invariably treated surgically. (2) For chronic MR, repair or replacement of the mitral valve is the current standard of care when patients become symptomatic with severe MR, dilated ventricles, new-onset atrial fibrillation, or pulmonary hypertension. (2) However, to prevent downstream complications, mitral valve repair or replacement may also be considered prophylactically in select asymptomatic patients. For example, in patients with flail leaflets (see below), early mitral valve surgery in asymptomatic patients was found to be more beneficial than watchful waiting. (3) In fact, it has been estimated that flail mitral valve leaflets will invariably require surgery within 10 years of diagnosis. (4)

Chronic MR is etiologically heterogeneous. The 2 most common categories are degenerative (myxomatous/fibroelastic deficiency causing flail leaflets or floppy valve leading to ineffective coaptation) and functional. (5) In functional MR, the mitral apparatus is structurally normal but its geometry is distorted. This may occur with left ventricular dilatation, ventricular remodelling, or dyssynchronous contraction of the left and right ventricles as observed post–myocardial infarction or in cardiomyopathy. (6) Surgical treatment of functional MR is less well established. (7)

Surgical correction of MR may employ valve repair or replacement (with or without preservation of the mitral apparatus). Surgical mitral valve repair is the treatment of choice in most cases because it preserves the mitral apparatus and competence, and does not require the lifelong anticoagulation needed if the valve were replaced with a prosthetic device. (8) Similar rates of reoperation (a postsurgical complication) have been observed following both valve repair and replacement, that is, about 20% over 19 years. (9) Despite the availability of the valvular repair option, the risk of undertaking the surgery may be too high or prohibitive; the Society of Thoracic Surgeons (STS) states that the risk of mortality is ≥ 8%. (10) In general this translates into a subgroup of symptomatic patients who are elderly, are frail, have several comorbidities, and have a history of previous cardiac procedures and/or hemodynamic instability. One meta-analysis has reported a pooled 5-year survival rate of 23% (95% confidence interval [CI], 12%–39%) in octogenarians undergoing surgical mitral valve repair. (11) As per another categorization, the risk of surgery is highest in those with a left ventricular (LV) ejection fraction < 0.6, an LV end-systolic dimension > 40 mm, or New York Heart Association (NYHA) class III or IV symptoms. (2)

As alternatives to surgery, various mitral valve repair procedures that employ a percutaneous approach are under investigation, such as edge-to-edge leaflet repair and indirect and direct annuloplasties. (5) In theory, percutaneous mitral valve repair would decrease the short-term risk of operative mortality and morbidity while increasing survival rates and quality of life.

Prevalence and Incidence

With prevalence rates as high as 9% in the elderly population aged ≥ 75 years, MR is the most common valvular abnormality observed in population-based studies. (12) About 50% of cases of isolated chronic MR may be of the severe symptomatic grade 3+/4+ variety, of which 49% may be at prohibitive risk for surgery. (13) By another estimation, many of the 500,000 North American patients with chronic MR may be poor candidates for mitral valve surgery. (14)

Technology/Technique

The mitral valve clip is one percutaneous mitral valve repair option, first used in humans in 2003. (15) The mitral valve clip system consists of the clip device, a delivery system, and a catheter guide through which the delivery system is introduced trans-septally into the left atrium toward the mitral valve through a femoral venous approach. Mitral valve clips have been implanted in more than 10,000 patients. (7) The clip device is conceptually based upon the commonly employed surgical edge-to-edge Alfieri technique of suturing the opposing leaflets in the middle, thereby creating 2 orifices for blood flow across the mitral opening. (16; 17)

Regulatory Status

In October 2013, the US Food and Drug Administration (FDA) approved mitral valve clips for use in patients with symptomatic MR (≥ 3+) of degenerative etiology who are at prohibitive risk for mitral valve surgery. (18) Clinical studies that supported this approval were these:

Endovascular Valve Edge-to-Edge Repair Study (EVEREST) I, a feasibility study
EVEREST II, a randomized controlled trial (RCT)
EVEREST II High Risk Registry study, a single arm-registry study
Real World Expanded Multi-center Study of the MitraClip System (REALISM) High Risk and REALISM Non-High Risk, continued access registry studies

In Europe, since the CE Marking approval in 2008, the device has been used in German and Italian centres primarily in patients with functional MR with low ejection fractions, and it has shown favourable acute and short-term outcomes based on longitudinal follow-up registry data (transcatheter mitral valve interventions [TRAMI] and ACCESS-Europe [ACCESS-EU] registries). (19;20) The Cardiovascular Outcomes Assessment of the MitraClip Therapy Percutaneous Therapy for High Surgical Risk Patients (COAPT) and the Randomized Study of the MitraClip Device in Heart Failure Patients With Clinically Significant Functional Mitral Regurgitation (RESHAPE-HF) are ongoing studies in high-risk patients with functional MR. (21)

There is only one mitral valve clip device approved for use in Canada for significant symptomatic degenerative mitral regurgitation (MR ≥ 3+) in patients who have been determined to be too high risk for mitral valve surgery by a heart team, including a cardiac surgeon, and in whom existing co-morbidities would not preclude the expected benefit from correction of the mitral regurgitation. The MitraClip system (licence number 93117) was approved on April 17, 2014, by Health Canada.

EVIDENCE-BASED ANALYSIS

Research Questions

For symptomatic patients at high or prohibitive surgical risk, what are the comparative effectiveness, harms, and cost-effectiveness of percutaneous mitral valve repair using a mitral valve clip?
What are the important population effect modifiers?

Research Methods

The methodological approach to evidence searching and synthesis conformed to the Cochrane Collaboration's methods guidance and followed an a priori protocol. As a priority, evidence was sought from the most recent and relevant existing systematic reviews (SRs) and health technology assessments (HTAs) if the documents included a broad and transparently reported search strategy, an appraisal of the validity of included studies, and a synthesis of the primary evidence aimed at minimizing bias. The search strategy had to report databases searched, search end dates, and screened identified studies using predefined eligibility criteria in order for the review to qualify as an SR and be assessed further for methodological rigour. If the evidence synthesis of available SRs/HTAs did not incorporate the risk of bias, but the searching and screening of literature were judged to be well conducted (i.e., reporting having searched at least 2 databases with 1 being MEDLINE; search end dates; and more than 1 reviewer screening), we used the most recent SR to identify relevant primary studies. Subsequent bibliographical searches updated the original review search, followed by a de novo evidence synthesis of originally included and newly identified studies.

Literature Search

A literature search was performed on January 26, 2014, using Ovid MEDLINE, Ovid MEDLINE In-Process and Other Non-Indexed Citations, and Embase using the OVID platform, for studies published within the past 10 years. (Appendix 1 provides details of the literature search strategies). We also searched the Cochrane Database of Systematic Reviews, Database of Abstracts of Reviews of Effect (DARE) on Wiley, NHS EED, and Tufts CEA registry. A focused grey literature search was undertaken using the Turning Research Into Practice (TRIP) database and the World Health Organization (WHO) Trial Registry, which also searches www.clinicaltrials.gov. A subsequent search to identify primary studies was performed on February 7, 2014. We applied a filter to identify SRs and HTAs and a separate filter for economic studies. While screening titles or abstracts, a single reviewer included eligible records; a second reviewer rescreened excluded records as an additional consideration. Full texts of included records were obtained and screened by 2 reviewers. Discrepancies were resolved by consensus or by involving a third team member. Systematic reviews were screened in reverse chronological order to identify 1 or 2 reviews with reliable search or methodological quality. Further, we screened primary literature to update the searches of identified reviews-searches overlapped by 6 months. Economic literature was searched and screened separately. Along with our a priori outcomes of interest, record eligibility criteria are reported below. Note, we did not exclude studies by outcomes.

Inclusion Criteria

Population: symptomatic patients (NYHA class III or IV) with moderate or severe MR (3+ or 4+) of chronic functional or degenerative etiology who are at high or prohibitive risk for surgical mitral valve repair or replacement (i.e., those who have a high risk based on STS score, or elderly with patients who have undergone previous cardiac surgery)
Intervention: mitral valve clip
Comparator: current standards of care—that is, surgical mitral valve repair or replacement or conservative medical treatment
Study designs and other criteria: SRs, HTAs, RCTs, comparative observational studies (including registry data), and full economic evaluations (i.e., cost-effectiveness analyses, cost-utility analyses, and cost-benefit analyses) reported in the English language

Exclusion Criteria

Studies involving populations with rheumatic heart disease, NYHA class I or II symptoms, and acute MR

Outcomes of Interest

Procedure-related adverse events (30-day mortality, 30-day stroke, cardiac perforation, blood transfusion requirement)
1-year all-cause mortality
NYHA class at 12 months
Quality of life
Readmission rates during the first 12 months
Recurrent (grade 3+ or 4+) MR at 12 months
Total serious adverse events (SAEs)
Ventricular remodelling (absolute end-systolic volume or change in end-systolic volume at 12 months)
Incremental cost-effectiveness

Risk of Bias Assessment

Primary study risk of bias was judged using the Cochrane risk of bias tool for RCTs. Generic assessment was employed to judge the risk of selection bias, confounding, and information bias (for a hypothetical target trial) for observational studies, and Phillips checklist was used to evaluate the methodological rigour of primary economic evaluations. For outcomes that were to be graded, publication bias was planned to be investigated when there were ≥ 10 studies contributing data for an outcome, studies were of unequal sizes, there were no important clinical and methodological differences between smaller and larger studies, and quantitative results were reported with accompanying measures of dispersion.

Synthesis of Evidence

Given the heterogeneity in comparative evidence, no meta-analysis could be performed. For the synthesis of the economic literature, we first identified common methodological issues within studies. Each study was assessed through a 3-step process: initial assessment for validity, assessment of the overall study quality (use of the Phillips checklist), and assessment of the study's quality and pertinence to the decision question. The focus was on the validity of evidence addressing cost-effectiveness of percutaneous mitral valve repair using a mitral valve clip when compared with the current standards of care. We also attempted to identify optimal patient subpopulations. Exploration of effect modification, especially by MR etiology, was planned but was not possible because of few studies contributing evidence.

Quality of Evidence

Using GRADEprofiler (version 3.6), the quality of the body of evidence for each a priori important outcome was judged according to the Grading of Recommendations Assessment, Development, and Evaluation (GRADE) Working Group criteria as high, moderate, low, or very low. (22)

Study design was the first consideration; the starting assumption was that RCTs are high quality, whereas observational studies are low quality because of concerns about residual and unmeasured confounding, even when some adjustments for confounding were made. Five additional factors—risk of bias, inconsistency, indirectness, imprecision, and publication bias—were then taken into account. Limitations in these areas resulted in downgrading the quality of evidence. None of the 3 optional domains that may raise the quality of evidence could be invoked.

As stated by the GRADE Working Group, the final quality score can be interpreted using the following definitions:

High	High confidence in the effect estimate—the true effect lies close to the estimate of the effect
Moderate	Moderate confidence in the effect estimate—the true effect is likely to be close to the estimate of the effect, but may be substantially different
Low	Low confidence in the effect estimate—the true effect may be substantially different from the estimate of the effect
Very Low	Very low confidence in the effect estimate—the true effect is likely to be substantially different from the estimate of the effect

Study Design	Number of Eligible Studies	Cost-Effectiveness Studies
Randomized controlled trial	1	1
Comparative cohort design	4	1
Interrupted time series	1	0
Uncontrolled before-after study	2	0
Total	8	2

Study and Design, Funding	Population	Total Sample Size	Comparator	Risk of Bias	Mortality (1 y Unless Specified and Highlighted)	Quality of Evidence^a
EVEREST II RCT, (25) Abbott Vascular	MR etiology: mixed (majority degenerative) Surgical risk: low NYHA III/IV = 50% MR status: 3+/4+ EF > 25%	279	Surgery	High 21% of patients receiving a mitral valve clip had subsequent surgery Attrition bias 4-y data	16 events RR, 1.08 (95% CI, 0.39–3.02) ARD = 4 more per 1,000 (from 34 fewer to 113 more)	Very low
			4-y mortality: 41 events RR, 0.98 (95% CI, 0.54–1.77) ARD = 4 more per 1,000 (from 82 fewer to 137 more)	Very low
Paranskaya et al, Germany, cohort, (29) NR	MR etiology: mixed (majority degenerative) Surgical risk: low NYHA III/IV = 92% MR status: 3+/4+ EF > 45%	50	Surgery	High Prognostic imbalance and variable duration of follow-up	2 events with mitral valve clips (not estimable)	Very low
Ajello et al, Italy, cohort, as abstract only, (30) NR	MR etiology: functional Surgical risk: mixed NYHA III/IV = 62% MR status: 3+/4+ (92%) EF mean = 27%	160	Surgery and conservative management equally distributed	High Selection bias (subset of data from the 2 groups were analyzed) and confounding	Favours mitral valve clips Actuarial survival = 88.9 ± 3.5% vs 69.5 ± 7.3% (P = 0.002)	Very low
			2-y mortality: Favours mitral valve clips Actuarial survival = 80.2 ± 5.2% vs 57.0 ± 8.1% (P = 0.002)	Very low
Buzzatti et al, before-after (predicted vs observed mortality), abstract only, (32) NR	MR etiology: mixed (70% functional) Surgical risk: unclear NYHA III/IV = 78% MR status: 3+/4+ (NR) EF mean = 36%	135	SHFM risk prediction (assumed a proxy for conservative management)	Unclear Some concerns about completeness of data and validity of SHFM risk prediction	Favours mitral valve clips for functional MR^b RD, 9.1% (95% CI, 3.1–15.1) Degenerative MR, RD = 11.1% (95% CI, −19.7 to 41.9)	Very low
			2-y mortality: Favours mitral valve clips for functional MR^b RD, 12.1% (95% CI, 3.1–15.1) Degenerative MR, RD, 22.2% (95% CI, −24.8 to 69.2)	Very low
EVEREST II HRS, cohort, (27) Abbott Vascular	MR etiology: mixed (58% functional) Surgical risk: high (some uncertainty) NYHA III/IV ≥ 80% MR status: 3+/4+ EF mean = 54% (excluded < 20%)	114	Conservative management and surgery (14%)	High Selection bias (post hoc control group) and confounding by indication	Favours mitral valve clips 35 events RR, 0.55 (95% CI, 0.32–0.94)	Very low
Schau and Neuss et al, Germany, before-after (predicted vs observed mortality), abstract only, (31;61) NR	MR etiology: mixed (73% functional) Surgical risk: high NYHA III/IV MR status: 3+/4+ EF: NR Duration of follow-up NR but probably > 6 mo	155	SHFM risk prediction (assumed a proxy for conservative management)	Unclear Some concerns about completeness of data and validity of SHFM risk prediction	Qualitative results showing no difference in predicted vs observed survival rate	Very low

Quality Assessment						No. of Patients			Effect	Quality
No. of Studies	Design	Risk of Bias	Inconsistency	Indirectness	Imprecision	Other Considerations	Mitral Valve Clip	Surgery	Relative (95% CI)	Quality	Absolute
Quality of life (physical component)^a
1 (25)	Randomized trial	Very serious^b	No serious inconsistency	Very serious^c	No serious imprecision	None	132	60	–	MD 0 higher (3.1 lower to 3.1 higher)	⊕ VERY LOW
Quality of life (mental component)^a
1 (25)	Randomized trial	Very serious^b	No serious inconsistency	Very serious^c	No serious imprecision	None	133	60	–	MD 1.9 higher (−1.2 lower to 5.0 higher)	⊕ VERY LOW

No. of Studies	Total No. of Patients	Study Design	Applicability
2 (25;29)	329 (sample size range: 50–279)	1 randomized controlled trial 1 cohort	Low applicability based on surgical mortality risk

Quality Assessment					Results		Quality
No. of Studies	Design	Structural Concerns	Concerns with Data	Concerns with Consistency	Main Estimate	Sensitivity Analyses	Quality
1 (38)	Trial-based analysis of mitral valve clip versus surgery (1-y duration)	Serious	Serious	Serious	Mitral valve clip increased QALYs by 0.015 versus surgery and reduced costs by $2,200 USD per patient versus surgery Mitral valve clip dominated surgery being more efficacious and less costly	Results were sensitive to assumptions regarding the duration of the QOL benefit with mitral valve clips relative to surgery, the price of mitral valve clips, and the analyzed population. Using a European price of $26,200 USD per clip, the ICER was greater than $400,000 USD per QALY	Low
1 (7)	Markov model of the mitral valve clip versus medical management (5-y duration)	No serious structural concerns	Serious	Moderate	Mitral valve clips resulted in 1.22 incremental QALYs versus MM and incurred additional costs of £27,000 ICER for mitral valve clips versus MM = £22, 153/QALY	Results were most sensitive to the time horizon, the utility decrement associated with NYHA II, and the cost of the mitral valve clip procedure; however, the ICER did not exceed £30,000 per QALY except with a time horizon of 2 y	Moderate

1	(Mitraclip* or Mitralclip* or (mitra adj clip) or (mitral adj clip)).tw. (738)
2	(Alfieri and (technique? or surger* or surgical* or repair*)).tw. (163)
3	“Endovascular Valve Edge-to-Edge Repair”.tw. (34)
4	(“edge-to-edge” and (endovascular* or percutaneous*)).tw. (274)
5	or/1–4 (991)
6	Mitral Valve/su [Surgery] (14176)
7	Mitral Valve Insufficiency/su, th [Surgery, Therapy] (16357)
8	Mitral Valve Prolapse/su, th [Surgery, Therapy] (1667)
9	Mitral Valve Annuloplasty/ (1387)
10	((mitral adj2 (valv* or insufficien* or incompeten* or prolaps* or regurgitat)) and (surger or surgical* or repair* or angioplast* or annuloplast* or catheter* or prothes?s or prosthetic* or plication*)).tw. (38202)
11	or/6–10 (52526)
12	Surgical Instruments/ (33792)
13	(clip* or clamp*).tw. (197878)
14	Heart Valve Prosthesis Implantation/is, mt [Instrumentation, Methods] (5300)
15	Surgical Procedures, Minimally Invasive/is, mt [Instrumentation, Methods] (8411)
16	Cardiac Catheterization/is, mt [Instrumentation, Methods] (7494)
17	exp Endovascular Procedures/is, mt [Instrumentation, Methods] (27158)
18	exp Angioplasty/is, mt [Instrumentation, Methods] (16279)
19	Mitral Valve Annuloplasty/is, mt (203)
20	exp Suture Techniques/is, mt [Instrumentation, Methods] (3816)
21	or/12–20 (277090)
22	11 and 21 (3682)
23	5 or 22 (4287)
24	exp Animals/ not (exp Animals/ and Humans/) (7994187)
25	23 not 24 (4106)
26	limit 25 to systematic reviews [Limit not valid in Embase; records were retained] (1593)
27	meta analysis.pt. (43391)
28	exp meta-analysis as topic/ (24364)
29	(meta-analy* or metanaly* or metaanaly* or met analy* or integrative research or integrative review* or integrative overview* or research integration or research overview* or collaborative review*).tw. (140645)
30	(systematic review* or systematic overview* or evidence-based review* or evidence-based overview* or (evidence adj3 (review* or overview)) or meta-review or meta-overview* or metasynthes* or “review of reviews” or technology assessment* or HTA or HTAs).tw. (172486)
31	exp Technology assessment, biomedical/ (20762)
32	(cochrane or health technology assessment or evidence report).jw. (25663)
33	or/27–32 (326529)
34	25 and 33 (39)
35	26 or 34 (1600)
36	(comment or editorial or interview or letter or news).pt. (2758368)
37	35 not 36 (1564)
38	37 use prmz (55)
39	implantable clip/ (159)
40	(Mitraclip? or Mitralclip? or (mitra adj clip?) or (mitral adj clip?)).tw. (719)
41	(Alfieri and (technique? or surger* or surgical* or repair*)).tw. (163)
42	“Endovascular Valve Edge-to-Edge Repair”.tw. (34)
43	(“edge-to-edge” and (endovascular* or percutaneous*)).tw. (274)
44	or/39–43 (997)
45	mitral valve/su [Surgery] (14176)
46	mitral valve regurgitation/su, th [Surgery, Therapy] (16357)
47	mitral valve prolapse/su, th [Surgery, Therapy] (1667)
48	mitral annuloplasty/ (1471)
49	((mitral adj2 (valv* or insufficien* or incompeten* or prolaps* or regurgitat)) and (surger or surgical* or repair* or angioplast* or annuloplast* or catheter* or prothes?s or prosthetic* or plication*)).tw. (38202)
50	or/45–49 (52538)
51	exp clip/ (28609)
52	(clip* or clamp*).tw. (197878)
53	suturing method/ (29026)
54	(sutur* adj3 (method* or technique*)).tw. (12984)
55	or/51–54 (254732)
56	50 and 55 (2480)
57	44 or 56 (3047)
58	exp animals/ or exp animal experimentation/ or exp models animal/ or exp animal experiment/ or nonhuman/ or exp vertebrate/ (37404701)
59	exp humans/ or exp human experimentation/ or exp human experiment/ (28288098)
60	58 not 59 (9118152)
61	57 not 60 (2902)
62	limit 61 to “reviews (maximizes specificity)” (14)
63	meta-analysis/ (123159)
64	“systematic review”/ (68978)
65	“meta analysis (topic)”/ (11251)
66	(meta-analy* or metanaly* or metaanaly* or met analy* or integrative research or integrative review* or integrative overview* or research integration or research overview* or collaborative review*).tw. (140645)
67	(systematic review* or systematic overview* or evidence-based review* or evidence-based overview* or (evidence adj3 (review* or overview)) or meta-review or meta-overview* or metasynthes* or “review of reviews” or technology assessment* or HTA or HTAs).tw. (172486)
68	biomedical technology assessment/ (19668)
69	(cochrane or health technology assessment or evidence report).jw. (25663)
70	or/63–69 (359716)
71	61 and 70 (30)
72	62 or 71 (31)
73	72 use emez (19)
74	38 or 73 (74)
75	limit 74 to last 10 years (70)
76	remove duplicates from 75 (61) [TOTAL UNIQUE RESULTS]
77	76 use prmz (51) [MEDLINE UNIQUE RESULTS]
78	76 use emez (10) [EMBASE UNIQUE RESULTS]

#1	(Mitraclip* or Mitralclip* or (mitra next clip) or (mitral next clip)):ti,ab,kw 10
#2	(Alfieri and (technique* or surger* or surgical* or repair*)):ti,ab,kw 0
#3	“Endovascular Valve Edge-to-Edge Repair”:ti,ab,kw 3
#4	(“edge-to-edge” and (endovascular* or percutaneous*)):ti,ab,kw 4
#5	{or #1-#4} 10
#6	[mh “Mitral Valve”/su] 188
#7	[mh “Mitral Valve Insufficiency”/su,th] 108
#8	[mh “Mitral Valve Prolapse”/su,th] 7
#9	[mh “Mitral Valve Annuloplasty”] 10
#10	((mitral near/2 (valv* or insufficien* or incompeten* or prolaps* or regurgitat)) and (surger or surgical* or repair* or angioplast* or annuloplast* or catheter* or prothes?s or prosthetic* or plication*)):ti,ab,kw 488
#11	{or #6-#10} 541
#12	[mh “Surgical Instruments”] 596
#13	(clip* or clamp*):ti,ab,kw 3834
#14	[mh “Heart Valve Prosthesis Implantation”/is,mt] 137
#15	[mh “Surgical Procedures, Minimally Invasive”/is,mt] 5977
#16	[mh “Cardiac Catheterization”/is,mt] 723
#17	[mh “Endovascular Procedures”/is,mt] 1868
#18	[mh Angioplasty/is,mt] 1144
#19	[mh “Mitral Valve Annuloplasty”/is,mt] 6
#20	[mh “Suture Techniques”/is,mt] 143
#21	{or #12-#20} 10964
#22	#11 and #21 89
#23	#5 or #22 from 2004 to 2014 65

ID	Search Hits
#1	(Mitraclip* or Mitralclip* or (mitra next clip) or (mitral next clip)):ti,ab,kw 17
#2	(Alfieri and (technique* or surger* or surgical* or repair*)):ti,ab,kw 0
#3	“Endovascular Valve Edge-to-Edge Repair”:ti,ab,kw 6
#4	(“edge-to-edge” and (endovascular* or percutaneous*)):ti,ab,kw 7
#5	{or #1-#4} 17
#6	[mh “Mitral Valve”/su] 193
#7	[mh “Mitral Valve Insufficiency”/su,th] 108
#8	[mh “Mitral Valve Prolapse”/su,th] 7
#9	[mh “Mitral Valve Annuloplasty”] 12
#10	((mitral near/2 (valv* or insufficien* or incompeten* or prolaps* or regurgitat)) and (surger or surgical* or repair* or angioplast* or annuloplast* or catheter* or prothes?s or prosthetic* or plication*)):ti,ab,kw 550
#11	{or #6-#10} 603
#12	[mh “Surgical Instruments”] 604
#13	(clip* or clamp*):ti,ab,kw 4091
#14	[mh “Heart Valve Prosthesis Implantation”/is,mt] 148
#15	[mh “Surgical Procedures, Minimally Invasive”/is,mt] 6114
#16	[mh “Cardiac Catheterization”/is,mt] 731
#17	[mh “Endovascular Procedures”/is,mt] 1909
#18	[mh Angioplasty/is,mt] 1159
#19	[mh “Mitral Valve Annuloplasty”/is,mt] 7
#20	[mh “Suture Techniques”/is,mt] 147
#21	{or #12-#20} 11378
#22	#11 and #21 99
#23	#5 or #22 from 2013 to 2014 8

Quality Assessment						Summary of Findings
Participants (Studies) Follow-Up	Risk of Bias	Inconsistency	Indirectness	Imprecision	Publication Bias	Overall Quality of Evidence (GRADE)	Study Event Rates (%)		Relative Effect (95% CI)	Anticipated Absolute Effects
Participants (Studies) Follow-Up	Risk of Bias	Inconsistency	Indirectness	Imprecision	Publication Bias	Overall Quality of Evidence (GRADE)	With Surgery and/or Conservative Treatment^a	With Mitral Valve Clip	Relative Effect (95% CI)	Risk with Surgery and/or Conservative Treatment^a	Risk Difference with Mitral Valve Clip (95% CI)
1-Year Mortality, EVEREST II RCT (25)
270 (1 study)	Very serious^b	No serious inconsistency	Very serious^c	Very serious^d	Undetected	⊕⊖⊖⊖ VERY LOW^b,c,d due to risk of bias, indirectness, imprecision	5/89 (5.6)	11/181 (6.1)	RR 1.08 (0.39–3.02)	56 per 1,000	4 more per 1,000 (from 34 fewer to 113 more)
4-Year Mortality, EVEREST II RCT (25)
234 (1 study)	Very serious^e	No serious inconsistency	Very serious^c	Very serious^d	Undetected	⊕⊖⊖⊖ VERY LOW^c,d,e due to risk of bias, indirectness, imprecision	13/73 (17.8)	28/161 (17.4)	RR 0.98 (0.54–1.77)	178 per 1,000	4 fewer per 1,000 (from 82 fewer to 137 more)

Follow-Up Period	Specific Outcomes	Random Sequence Generation	Allocation Concealment	Blinding of Participants/Personnel	Blinding of Outcome Assessment	Incomplete Outcome Data	Selective Reporting	Other Bias (Confounding)
1 y	All other outcomes with data	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Low risk of bias^e	Low risk of bias	Unclear risk of biasⁱ	High risk of bias^j
1 y	Quality of life	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	High risk of bias^f	High risk of bias^h	Unclear risk of biasⁱ	High risk of bias^j
1 y	Total serious adverse events	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	High risk of bias^f	Unclear risk of bias	Unclear risk of biasⁱ	High risk of bias^j
1 y	NYHA III/IV, end-systolic volume	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Low risk of bias^e	High risk of bias^h	Unclear risk of biasⁱ	High risk of bias^j
4 y	All outcomes with data	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Low risk of bias^e	High risk of bias^h	Unclear risk of biasⁱ	High risk of bias^j
4 y	Postintervention surgery	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Low risk of bias^e	High risk of bias^h	Unclear risk of biasⁱ	Unclear risk of bias
30 d	Transfusion, stroke, and mortality	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Low risk of bias^e	Low risk of bias	Low risk of bias	Low risk of bias
30 d	Major adverse events	Low risk of bias^b	Low risk of bias^c	Low risk of bias^d	Unclear risk of bias^g	Low risk of bias	Unclear risk of biasⁱ	Low risk of bias

Authors, Year (Authors, Year of Companion)	Analysis Type	Subgroup	Study Design	Maximum Duration of Follow-Up	Total Sample Size	Funding	Conflict of Interest
Mauri et al, 2013 (25)	Main study	n/a	RCT	4y	279	Industry	Yes
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Companion	Patients with AF	RCT	1 y	279	Industry	Yes
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Companion	Patients without AF	RCT	1 y	279	Industry	Yes
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Companion	n/a	RCT	1 y	279	Industry	Yes

Authors, Year (Authors, Year of Companion)	NYHA Class III or IV	MR 3+ or 4+	Surgical Mortality Risk?^a	Elderly with History of Cardiac Surgery?	Type of MR	AF?	Pulmonary Hypertension?	Concomitant Severe RV Dysfunction at Baseline?	LVEF < 25% at Baseline?
Mauri et al, 2013 (25)	Not majority	Yes, majority	No	No	Mixed	NR	NR	NR	No
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Not majority	Yes, all	No	No	Mixed	Yes	NR	NR	No
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Not majority	Yes, all	No	No	Mixed	Yes	NR	NR	No
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Not majority	Yes, majority	No	No	Mixed	NR	NR	NR	No

Authors, Year (Authors, Year of Companion)	Domain	Outcome 1 (Prespecified)	Outcome 2 (Not Prespecified but Serves in Main Domain)	Definition of Outcome (How Measured and Categorized)	Qualitative Results	Duration of Observation/Time Point
Mauri et al, 2013 (25)	Mortality	1-y all-cause mortality	n/a	n/a	n/a	1 y
Mauri et al, 2013 (25)	Mortality	Other	Death	n/a	n/a	4 y
Mauri et al, 2013 (25)	Other	Other	Composite outcome	Freedom from death, MV surgery or reoperation, and MR 3+ or 4+	n/a	4 y
Mauri et al, 2013 (25)	Failure	Postintervention surgery	MV surgery or reoperation	n/a	n/a	4 y
Mauri et al, 2013 (25)	Failure	Other	MR 3+ or 4+ at follow-up	n/a	n/a	1 y
Mauri et al, 2013 (25)	Failure	Other	MR 3+ or 4+ at follow-up	n/a	n/a	4 y
Mauri et al, 2013 (25)	Function/QOL	NYHA class at 12 mo		n/a	n/a	1 y
Mauri et al, 2013 (25)	Function/QOL	Other	NYHA class III or IV at 4 y	n/a	n/a	4 y
Mauri et al, 2013 (25)	Other	Other	Composite outcome in patients with degenerative MR	n/a	n/a	4 y
Mauri et al, 2013 (25)	Other	Other	Composite outcome in patients with functional MR	n/a	n/a	4 y
Mauri et al, 2013 (25)	Ventricular remodelling	Ventricular remodelling (absolute end-systolic volume)	n/a	Left ventricular end-systolic volume	n/a	4 y
Mauri et al, 2013 (25)	Ventricular remodelling	Ventricular remodelling (absolute end-systolic volume)	n/a	Left ventricular end-systolic volume	n/a	1 y
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Mortality	Procedure-related AEs (30-d mortality)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Mortality	Procedure-related AEs (30-d mortality)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Stroke	Procedure-related AEs (30-d stroke)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Stroke	Procedure-related AEs (30-d stroke)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Harms	Procedure-related AEs (blood transfusion requirement)	n/a	≥ 2 units of blood	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Harms	Procedure-related AEs (blood transfusion requirement)	n/a	≥ 2 units of blood	n/a	30 d
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	Harms	Total major AEs	n/a	Composite outcome^a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Total major AEs	n/a	Composite outcome^a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Major SAEs	n/a	Composite outcome^a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Other	n/a	Any major AEs	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Total SAEs (probably FDA defined)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Mortality	Procedure-related AEs (30-d mortality)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Stroke	Procedure-related AEs (30-d stroke)	n/a	n/a	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Procedure-related AEs (blood transfusion requirement)	n/a	≥ 2 units of blood	n/a	30 d
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Function/QOL	QOL	Physical component summary	n/a	n/a	1 y
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Function/QOL	QOL	Mental component summary	n/a	n/a	1 y
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Ventricular remodelling	Ventricular remodelling (change in end-systolic volume at 12 mo)	n/a	Change from baseline	n/a	1 y
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	Harms	Overall conclusions about AEs	Renal failure, septicemia, MI, gastrointestinal complications requiring surgery	Renal^b	30 d

Authors, Year	Intervention				Comparator				Between-Group		Comparative Estimates
Authors, Year	No. of Events	No. Analyzed	Mean	SD	No. of Events	No. Analyzed	Mean	SD	Estimate	SD	LCI	UCI	P Value
Mauri et al, 2013 (25)	11	181			5	89			n/a				1
Mauri et al, 2013 (25)	28	161			13	73			n/a				0.914
Mauri et al, 2013 (25)	64	161			39	73			n/a				0.07
Mauri et al, 2013 (25)	40	161			4	73			n/a				<0.001
Mauri et al, 2013 (25)	38	181			18	89			n/a				1
Mauri et al, 2013 (25)	35	161			18	73			n/a				0.745
Mauri et al, 2013 (25)	3	151			9	67			n/a
Mauri et al, 2013 (25)	6	105			3	48			n/a
Mauri et al, 2013 (25)	49	117			34	51			−25		−41	−9
Mauri et al, 2013 (25)	15	44			5	22			11		−11	34
Mauri et al, 2013 (25)		94	54.46	24.2		41	48.93	27.9	5.5		−4	15	0.247
Mauri et al, 2013 (25)		144	57.54	24.04		66	55.74	31.4	1.8		−6	9.6	0.68
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	1	45			1	27			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	1	130			1	62			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	1	45			1	27			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	1	130			1	62			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	8	45			17	27			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	15	130			23	62			n/a
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))	10	45			18	27			n/a
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	16	130			25	62			n/a
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	27	180			45	94			n/a				<0.001
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	9	180			9	94			n/a				0.23
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	93	184			46	95			n/a
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	2	180			2	94			n/a				0.89
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	2	180			2	94			n/a				0.89
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))	24	180			42	94			n/a				<0.001
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))		132	4.4	9.8		60	4.4	10.4	n/a				0.002
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))		133	5.7	9.9		60	3.8	10.3	n/a				0.006
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))		144	−5.5	14.5		66	−5.6	21	n/a
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))

Authors, Year of Publication	Outcome Category	Specific Outcomes	Selection Bias	Explanation	Confounding Bias	Explanation	Information/Measurement Bias	Explanation
Conradi et al, 2013 (26)	All	Mortality 30 d, 3+/4+ MR, postintervention surgery, stroke 30 d, transfusion, and NYHA class III/IV	Low risk of bias		High risk of bias	See footnote^a	Low risk of bias
Conradi et al, 2013 (26)	Specific	Mortality at 6 mo	Low risk of bias		Low risk of bias	See footnote^b	Low risk of bias
Conradi et al, 2013 (28)	All	30-day mortality; redo cardiac surgery	High risk of bias	See footnote^c	High risk of bias	See footnote^d	Unclear risk of bias	See footnote^e
Paranskaya et al, 2013 (29)	All	Postintervention surgery; MR3+ last follow-up or 1 y; NYHA class at follow-up; stroke at 30 d; important procedure-related complications (various adverse clinical events); mortality at 30 d; mortality at last follow-up or 1 y; LVESV at last follow-up (mL)	High risk of bias	Patients had variable lengths of follow-up, while most data were analyzed as proportions	High risk of bias	Unadjusted estimates with major prognostic imbalance between groups	Low risk of bias
Buzzatti et al, 2013 (32)	All	Predicted minus observed mortality in functional and degenerative MR population	Unclear risk of bias	Some concerns about missing data	Unclear risk of bias	See footnote^f	Unclear risk of bias	Reliability and completeness of patient data for the model not reported
Ajello et al, 2013 (30)	All	NYHA at last follow-up; MR ≥ 3+ at last follow-up	High risk of bias	See footnote^g	High risk of bias	See footnote^h	Low risk of bias
Schau et al, 2013 (31)	All	Survival	Unclear risk of bias	Insufficient data	Unclear risk of bias	See footnoteⁱ	Unclear risk of bias	Insufficient data
Whitlow et al, 2012 (27)	All	All-cause mortality; 30-d mortality	High risk of bias	See footnote^j	High risk of bias	See footnote^k	High risk of bias	See footnote^l

Quality Criteria	Questions for Critical Appraisal	Response	Comments
S1	Is there a clear statement of the decision problem?	Yes
	Is the objective of the evaluation and model specified and consistent with the stated decision problem?	Yes
	Is the primary decision-maker specified?	No
S2	Is the perspective of the model stated clearly?	No
	Are the model inputs consistent with the stated perspective?	Unclear
	Has the scope of the model been stated and justified?	No
	Are the outcomes of the model consistent with the perspective, scope, and overall objective of the model?	Unclear
S3	Has the evidence regarding the model structure been described?	No
	Is the structure of the model consistent with a coherent theory of the health condition under evaluation?	Unclear
	Are the sources of data used to develop the structure of the model specified?	Yes/no	Unclear where costs for rehabilitation and long-term care are sourced
S4	Are the structural assumptions transparent and justified?	n/a	Not a model
	Are the structural assumptions reasonable given the overall objective, perspective, and scope of the model?	n/a	Not a model
S5	Is there a clear definition of the options under evaluations?	Yes
	Is there justification for the exclusion of feasible options?	No	Not discussed
S6	Is the chosen model type appropriate given the decision problem and specified causal relationship within the model?	n/a	Not a model
S7	Is the time horizon of the model sufficient to reflect all important differences between options?	n/a	Not a model
	Are the time horizon of the model, the duration of treatment, and the duration of treatment effect described and justified?	n/a	Not a model
S8	Do the disease states (state transition model) or the pathways (decision tree model) reflect the underlying biological process of the disease in question and the impact of interventions?	n/a	Not a model
S9	Is the cycle length defined and justified in terms of the natural history of disease?	n/a	Not a model
D1	Are the data identification methods transparent and appropriate given the objectives of the model?	Yes/no	Reporting lacks clarity as only available in abstract
	Where choices have been made between data sources, are these justified appropriately?	No
	Has particular attention been paid to identifying data for the important parameters in the model?	n/a	Not a model
	Has the process of selecting key parameters been justified and systematic methods used to identify the most appropriate data?	Unclear
	Has the quality of the data been assessed appropriately?	Unclear
	Where expert opinion has been used, are the methods described and justified?	Unclear
D2	Is the premodel data analysis methodology based on justifiable statistical and epidemiological techniques?	Unclear
D2a	Is the choice of baseline data described and justified?	n/a	Not a model
	Are transition probabilities calculated appropriately?	n/a	Not a model
	Has a half-cycle correction been applied to both cost and outcome?	n/a	Not a model
	If not, has this omission been justified?	n/a	Not a model
D2b	If relative treatment effects have been derived from trial data, have they been synthesized using appropriate techniques?	Unclear	Unclear how quality of life from 1 mo through to 12 mo was calculated
	Have the methods and assumptions to extrapolate short-term results to final outcomes been documented and justified?	Unclear	See above
	Have alternative extrapolation assumptions been explored through sensitivity analysis?	n/a	Not extrapolation of data
	Have alternative assumptions regarding the continuing effect of treatment been explored through sensitivity analysis?	n/a	Not extrapolation of data
D2c	Are the utilities incorporated into the model appropriate?	Unclear	See above
	Is the source for the utility weights referenced?	Yes	Derived from clinical trial
	Are the methods of derivation for the utility weights justified?	Unclear
D3	Have all data incorporated into the model been described and referenced in sufficient detail?	No	Brief summary within abstract
	Has the use of mutually inconsistent data been justified (i.e., are assumptions and choices appropriate)?	Unclear
	Is the process of data incorporation transparent?	No
	If data have been incorporated as distributions, has the choice of distribution for each parameter been described and justified?	n/a
	If data have been incorporated as distributions, is it clear that second-order uncertainty is reflected?	n/a
D4	Have the 4 principal types of uncertainty been addressed?	No
	If not, has the omission of particular forms of uncertainty been justified?	No
D4a	Have methodological uncertainties been addressed by running alternative versions of the model with different methodological assumptions?	n/a
D4b	Is there evidence that structural uncertainties have been addressed via sensitivity analysis?	n/a
D4c	Has heterogeneity been dealt with by running the model separately for different subgroups?	n/a
D4d	Are the methods of assessment of parameter uncertainty appropriate?	Unclear	Limited reporting of sensitivity analyses
	If data are incorporated as point estimates, are the ranges used for sensitivity analysis stated clearly and justified?	n/a
C1	Is there evidence that the mathematical logic of the model has been tested thoroughly before use?	n/a
C2	Are the conclusions valid given the data presented?	Yes
	Are any counterintuitive results from the model explained and justified?	No	RCT favoured surgical treatment
	If the model has been calibrated against independent data, have any differences been explained and justified?	n/a
	Have the results of the model been compared with those of previous models and any differences in results explained?	No	Results not put into context with previous literature

Quality Criteria	Questions for Critical Appraisal	Response	Comments
S1	Is there a clear statement of the decision problem?	Yes	“As patients within the EVEREST II HRS trial were old and infirm and the cost of the implant is incurred on day 1, it is important to know if these individuals incur enough benefit to overcome the initial expenditure”
	Is the objective of the evaluation and model specified and consistent with the stated decision problem?	Yes	“To assess the cost-effectiveness of the MitraClip therapy compared to medical management (MM) in patients with severe MR, for whom surgery is not an option due to high operative risk, with the primary data source being the EVEREST II High Risk Study”
	Is the primary decision-maker specified?	Yes	UK National Health Service
S2	Is the perspective of the model stated clearly?	Yes	Health care system
	Are the model inputs consistent with the stated perspective?	Yes
	Has the scope of the model been stated and justified?	Yes/no	Presurgical costs were not incorporated within the model. The authors justify this based on the fact that the difference in workup costs between the 2 treatment arms would have to be greater than £9,600 to alter the cost-effectiveness decision
	Are the outcomes of the model consistent with the perspective, scope, and overall objective of the model?	Yes	Cost per QALY gained
S3	Has the evidence regarding the model structure been described?	Yes	Underlying assumptions validated by advisory panel and experienced clinicians
	Is the structure of the model consistent with a coherent theory of the health condition under evaluation?	Yes	Similar to other models within this area
	Are the sources of data used to develop the structure of the model specified?	Yes	Efficacy: from EVEREST HRS II for mitral valve clip; from historical matched controls for the medical management group
			Costs: standard references for drug costs, hospital costs, adverse event costs, valve cost
			Utilities: from published literature sources
S4	Are the structural assumptions transparent and justified?	Yes/no	Mortality long term is not transparent—it appears that the difference in mortality between the 2 groups is assumed to continue for the duration of the model, although the clinical trial data for mitral valve clip are for 12 mo
	Are the structural assumptions reasonable given the overall objective, perspective, and scope of the model?	Yes/no	Concerns regarding long-term mortality differences, as detailed above
S5	Is there a clear definition of the options under evaluation?	No	Mitral valve clip versus MM; MM is not clearly defined
	Is there justification for the exclusion of feasible options?	Yes	Patients are not eligible for surgery due to high risk of mortality
S6	Is the chosen model type appropriate given the decision problem and specified causal relationship within the model?	Yes	Markov model, 2 segments, short-term 30-d horizon with 1-d cycle, long-term 5-y horizon with 1-mo cycles
S7	Is the time horizon of the model sufficient to reflect all important differences between options?	Yes	Base-case time horizon is 5 y; other durations tested in sensitivity analyses, including 2 y and 10 y
	Are the time horizon of the model, the duration of treatment, and the duration of treatment effect described and justified?	Yes/no	It is unclear how long-term survival was derived and if differences between treatments were assumed; “the cumulative survival estimates were extrapolated into the future and used to derive daily or monthly transition probabilities using standard formulas during all model cycles”
S8	Do the disease states (state transition model) or the pathways (decision tree model) reflect the underlying biological process of the disease in question and the impact of interventions?	Yes
S9	Is the cycle length defined and justified in terms of the natural history of disease?	Yes	Daily cycle length during the 30-d short-term model, and 1-mo cycle lengths during the long-term 5-y period of the model
D1	Are the data identification methods transparent and appropriate given the objectives of the model?	Yes	Efficacy for mitral value is from EVEREST II HRS; for MM it is from a historical cohort
			Costs—standard sources
			Mortality—long-term follow-up from EVEREST II HRS; assumptions regarding medical management tested in sensitivity analyses
			Utilities—published literature
	Where choices have been made between data sources, are these justified appropriately?	Yes
	Has particular attention been paid to identifying data for the important parameters in the model?	Yes
	Has the process of selecting key parameters been justified and systematic methods used to identify the most appropriate data?	Yes/no	The application of 2-y NYHA class distribution data from the EVEREST II study that was not in patients ineligible for surgery may not be justified. Concerns regarding the assumptions in the MM group specific to long-term mortality, NYHA classification, and comparability of patients
	Has the quality of the data been assessed appropriately?	Yes/no	See above
	Where expert opinion has been used, are the methods described and justified?	Yes
D2	Is the premodel data analysis methodology based on justifiable statistical and epidemiological techniques?	Yes/no	The equivalence of the treatment group and the historical control may be of concern; 58% of historic controls did not meet 1 of the inclusion criteria for the treatment group
D2a	Is the choice of baseline data described and justified?	Yes	Assumed to have the same mix of NYHA classification
	Are transition probabilities calculated appropriately?	Unclear	Difficult to assess
	Has a half-cycle correction been applied to both cost and outcome?	No
	If not, has this omission been justified?	No
D2b	If relative treatment effects have been derived from trial data, have they been synthesized using appropriate techniques?	No	Treatment effects were derived from EVEREST HRS II. The patients were matched with historical controls. The equivalence of the groups may be a concern
	Have the methods and assumptions to extrapolate short-term results to final outcomes been documented and justified?	Unclear	Methods for long-term mortality are unclear
	Have alternative extrapolation assumptions been explored through sensitivity analysis?	Yes/no	In some cases
	Have alternative assumptions regarding the continuing effect of treatment been explored through sensitivity analysis?	No
D2c	Are the utilities incorporated into the model appropriate?	Yes
	Is the source for the utility weights referenced?	Yes/no	Utility values are not sourced directly from patients undergoing procedure; rather, they are derived from published literature that reports utility decrements associated with NYHA classifications, decrements associated with ICU and non-ICU stays, treatment-related adverse events, and mitral valve regurgitation
			For MR, the article references 2 studies, both measured quality of life: 1 in patients with cardiomyopathy and 1 in patients undergoing mitral valve repair or replacement; however, neither study reported the disutility value used within this analysis
	Are the methods of derivation for the utility weights justified?	Yes/no	The derivation of the disutility associated with mitral regurgitation is unclear
			A disutility appears to have been applied for both MR and NYHA class, which may lead to an overestimation of the disutility
D3	Have all data incorporated into the model been described and referenced in sufficient detail?	Yes/no	Some details regarding the transition probabilities, mortality, and utilities are unclear
	Has the use of mutually inconsistent data been justified (i.e., are assumptions and choices appropriate)?	n/a
	Is the process of data incorporation transparent?	Yes/no	In most cases
	If data have been incorporated as distributions, has the choice of distribution for each parameter been described and justified?	No	Distributional assumptions were not reported
	If data have been incorporated as distributions, is it clear that second-order uncertainty is reflected?	Yes	Results of PSA are presented as a cost-effectiveness acceptability curve
D4	Have the 4 principal types of uncertainty been addressed?	No	See details below under D4a to D4d
	If not, has the omission of particular forms of uncertainty been justified?	No	See details below under D4a to D4d
D4a	Have methodological uncertainties been addressed by running alternative versions of the model with different methodological assumptions?	Yes	Different discount rates and time horizons have been incorporated. Some alternative values for health deficits have been considered; however, the methods are unclear
D4b	Is there evidence that structural uncertainties have been addressed via sensitivity analysis?	Yes/unclear	Assumptions regarding long-term mortality with treatment do not appear to have been tested in sensitivity analyses. Assumptions regarding long-term mortality with MM were examined in sensitivity analyses, as were different utility valuations and varying decrements
D4c	Has heterogeneity been dealt with by running the model separately for different subgroups?	No
D4d	Are the methods of assessment of parameter uncertainty appropriate?	Unclear	Assumptions regarding parameter distributions within the PSA are unclear
	If data are incorporated as point estimates, are the ranges used for sensitivity analysis stated clearly and justified?	Yes Unclear	Details of the ranges within deterministic sensitivity analyses are provided
C1	Is there evidence that the mathematical logic of the model has been tested thoroughly before use?	Unclear
C2	Are the conclusions valid given the data presented?	Yes
	Are any counterintuitive results from the model explained and justified?	Yes
	If the model has been calibrated against independent data, have any differences been explained and justified?	No
	Have the results of the model been compared with those of previous models and any differences in results explained?	Yes

Authors, Year (Authors, Year of Companion)	Exclusions	Is Study Applicable to My Population of Interest?	Intervention	Comparator
Mauri et al, 2013 (25)	MV area < 4.0 cm²; severe leaflet/annular calcification; flail width ≥ 15 mm; flail gap ≥ 10 mm; pts with functional etiology having coaptation depth > 11 mm below the annulus or coaptation length < 2 mm	Low applicability	MV clip	Surgery
Mauri et al, 2013 (25) (Herrmann et al, 2012 (34))		Low applicability	MV clip	Surgery
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))		Low applicability	MV clip	Surgery
Mauri et al, 2013 (25) (Feldman et al, 2011 (36))		Low applicability	MV clip	Surgery

Authors (Year)	Study Design	Single/Multicentre	Country Status	Maximum Duration of Follow-Up	Sample Size	Funding
Whitlow et al, 2012 (27)	Comparative observational cohort	Multiple	Developed	1 y	150	Industry
Conradi et al, 2013 (26)	Comparative observational cohort	Single	Developed	6 mo	171	Industry
Paranskaya et al, 2013 (29)	Comparative observational cohort	Single	Developed	32 d to 3.8 y (median, 1.3 y)	50	NR
Buzzatti et al, 2013 (32)	Predicted counterfactual comparative study	Single	NR	2 y	144	NR
Ajello et al, 2013 (30)	Comparative observational cohort	Single	NR	October 2008–February 2013	252	NR
Schau et al, 2013 (31)	Predicted counterfactual comparative study	NR	NR	March 2009 and November 2012	155	NR
Conradi et al, 2013 (28)	Before-after (time series)	Single	Developed	6 mo	446	NR

Authors (Year)	Exclusions	Is Study Applicable to My Population of Interest?	Intervention	Comparator
Whitlow et al, 2012 (27)	See footnote^a	High applicability	MV clip	Other (explain)
Conradi et al, 2013 (26)	NR	Mix of low to high applicability	MV clip	Surgery
Paranskaya et al, 2013 (29)	Mitral stenosis and endocarditis	Low applicability	MV clip	Surgery
Buzzatti et al, 2013 (32)	NR	Unclear applicability	MV clip	SHFM risk prediction model
Ajello et al, 2013 (30)	NR	Mix of low to high applicability	MV clip	Other (explain)
Schau et al, 2013 (31)	NR	Unclear applicability	MV clip	SHFM risk prediction model
Conradi et al, 2013 (28)	NR	Unclear applicability	Before MV clip procedure	After MV clip procedure

Authors (Year)	Intervention				Comparator				Between-Group		Comparative Estimates
Authors (Year)	No. of Events	No. Analyzed	Mean	SD	No. of Events	No. Analyzed	Mean	SD	Estimate	SD	LCL	UCL	P Value
Whitlow et al, 2012 (27)	6	78			3	36			n/a
Whitlow et al, 2012 (27)	19	78			16	36			n/a
Conradi et al, 2013 (26)	4	95			2	76			n/a
Conradi et al, 2013 (26)	8	95			4	76			n/a				NS
Conradi et al, 2013 (26)	4	95			1	76			n/a
Conradi et al, 2013 (26)	3	95			0	76			n/a
Conradi et al, 2013 (26)	1	95			0	76			n/a
Conradi et al, 2013 (26)		95	0.5	1.2		76	1.4	1.7	n/a
Conradi et al, 2013 (26)	29	95			13	76			n/a
Conradi et al, 2013 (26)	1	95			0	76		1	n/a
Paranskaya et al, 2013 (29)	0	24			0	26			n/a
Paranskaya et al, 2013 (29)	2	24			0	26			n/a
Paranskaya et al, 2013 (29)	0	24			2	26			n/a				0.29
Paranskaya et al, 2013 (29)	3	24			1	26			n/a
Paranskaya et al, 2013 (29)	0	24			1	26			n/a				0.65
Paranskaya et al, 2013 (29)	1	24			2	26			n/a
Paranskaya et al, 2013 (29)	8	24			4	26			n/a				0.51
Paranskaya et al, 2013 (29)		24	55.4	19		26	57	15.6	n/a				0.79
Paranskaya et al, 2013 (29)
Buzzatti et al, 2013 (32)									9.1				0
Buzzatti et al, 2013 (32)									12.1				<0.0001
Buzzatti et al, 2013 (32)
Ajello et al, 2013 (30)	10	105			37	112							0.03
Ajello et al, 2013 (30)	19	105			41	112			n/a				0.46
Ajello et al, 2013 (30)		105	80.2 ± 5.2			112	57.0 ± 8.1		n/a				0
Ajello et al, 2013 (30)		105	88.9 ± 3.5			112	69.5 ± 7.3		n/a				0
Schau et al, 2013 (31)
Conradi et al, 2013 (28)	14	194			11	252			n/a				0.22
Conradi et al, 2013 (28)									0.52				<0.01

Ref ID	Reviewer 1	Reviewer 2	Consensus	Additional Notes
75	I	I	I
80	I	I	I
103	E	E	Does not evaluate benefits or harms of mitral valve clip or percutaneous MR edge-to-edge repair (1b)
104	I	I	I	Also AF effect modification
106	I	I	I	Postprocedure repair rates (serious adverse event?)
517	E	E	Does not evaluate benefits or harms of mitral valve clip or percutaneous MR edge-to-edge repair, but investigates CARILLON Mitral Contour System (1b)
522	I	I	I
556	I	E	2 groups, but different populations as descriptive results (1c)
610	I	I	I
626	E	E	Descriptive (1c)
633	E	E	1c
1,006	E	E	1c
1,007	E	E	1c
1,016	E	E	1c
1,057	E	I	I	Transparently report post hoc inclusion of this design that examines health services/program improvement
1,121	I	E	1c
1,175	I	I	I
1,204	E	E	1c
1,270	U	I	I
1,271	U	I	I	Transparently report this novel design for a counterfactual comparator
1,282			Full text could not be retrieved
1,292	I	I	I
1,326	I	I	I
1,338	E	E	1c
1,353	I	I	I
1,377	E	E	1c
1,381	E	E	1c
1,389	E	E	1c
1,414	E	E	1c
1,434	E	E	1c
1,500	E	E	1c	Consider methods for QA assessment
1,502	E	E	1c	Consider
1,503	E	E	1c	Consider
1,504	E	E	1c	Consider
1,505	E	E	1c	Consider
1,506	E	E	1c	Consider

PERMALINK

Mitral Valve Clip for Treatment of Mitral Regurgitation: An Evidence-Based Analysis

Mohammed T Ansari, MBBS, MMEDSC, MPHIL

Nadera Ahmadzai, MD, MPH, MSc

Kathryn Coyle, BSCPHARM, MSc

Doug Coyle, MA, MSc, PhD

David Moher, PhD

Abstract

Background

Objective

Data Sources

Review Methods

Results

Limitations

Conclusions

PLAIN LANGUAGE SUMMARY

BACKGROUND

Objective of Analysis

Clinical Need and Target Population

Description of Condition

Prevalence and Incidence

Technology/Technique

Regulatory Status

EVIDENCE-BASED ANALYSIS

Research Questions

Research Methods

Literature Search

Inclusion Criteria

Exclusion Criteria

Outcomes of Interest

Risk of Bias Assessment

Synthesis of Evidence

Quality of Evidence

Results of Evidence-Based Analysis

Search Yield

Figure 1: PRISMA Flow Diagram for the Effectiveness and Harms of Mitral Valve Clip, and Economic Review.

Study Selection

Records Addressing Effectiveness and Harms of Mitral Valve Clips

Records Addressing Economic Evaluation of Mitral Valve Clips

Summary of Included Studies and Methodological Rigour

Table 1:

Table 2:

Post Hoc Changes or Decisions

Comparative Effectiveness and Harms

Mortality

Table 3:

Table 4:

Quality of Life

Table 5:

Patients in NYHA Class III/IV

One-Year Follow-Up

Table 6:

Table 7:

Follow-Up Greater Than 1 Year

Table 8:

Readmission Rates

Recurrent Mitral Regurgitation 3+/4+

Table 9:

Table 10:

Ventricular Remodelling (End-Systolic Volume)

Table 11:

Table 12:

Serious Adverse Events (Total)

Table 13:

Procedure-Related Adverse Events

30-Day Mortality

Table 14:

Table 15:

30-Day Stroke and Cardiac Perforation

Periprocedural Blood Transfusion

Table 16:

Other Important Adverse Events (Post Hoc)

Cost-Effectiveness of Mitral Valve Clips

Table 17:

Table 18:

CONCLUSIONS

ACKNOWLEDGEMENTS

APPENDICES

Appendix 1: Literature Search Strategies

Appendix 2: Evidence Quality Assessment