Systematic Review of Cost-Effectiveness Analyses Comparing Open and Minimally Invasive Lumbar Spinal Surgery

Kelechi Eseonu; Uche Oduoza; Mohamed Monem; Mohamed Tahir

doi:10.14444/8297

Article Figures & Data

Figures

Tables
Supplementary Materials

Download figure
Open in new tab
Download powerpoint
Figure 1
Study selection flow chart of the results of systematic literature review for identification of included studies.
Download figure
Open in new tab
Download powerpoint
Figure 2
UK health care spending as a percentage of gross domestic product.
Download figure
Open in new tab
Download powerpoint
Figure 3
Number of clinical-effectiveness studies in the lumbar spine indexed per year in PubMed (1992–current).

View popup

Table 1

Summary of inclusion and exclusion criteria.

Variable(s)	Inclusion Criteria	Exclusion Criteria
Patient Factors	Adult patients with back and/or leg pain who underwent surgery for the following degenerative conditions: Lumbar stenosis Lumbar radiculopathy Spondylolisthesis	Aged <18 y Tumor Infection Cervical or thoracic spinal surgery Fracture/trauma Revision surgery Scoliosis Spinal injections
MAS^{^a}	Single-level or multilevel posterior MAS including Laminectomy (any type) Microdiscectomy TLIF PLIF	Cervical or thoracic surgery Anterior or lateral lumbar surgery Computer-assisted or robotic surgery Kyphoplasty or vertebroplasty
Conventional (open) surgery	Open posterior spinal procedures including Laminectomy (any type) Microdiscectomy TLIF PLIF	Comparisons of 2 MAS procedures Cervical or thoracic surgery Anterior or lateral lumbar surgery
Outcomes	Costs analysis only Incremental cost-effectiveness ratio (or similar cost-effectiveness metric) Both open and MAS approach both analyzed for cost and/or outcomes	Differences in clinical outcomes only without cost analysis Utilities only Procedural costs not included
Study type	Full economic studies (cost-utility, cost-effectiveness, cost-benefit, cost-minimization) Cost studies (retrospective cohorts, nonrandomized prospective cohorts, decision model analyses)	Studies with <10 patients per treatment arm Systematic reviews
Publication type	Studies published in English Peer-reviewed journals	Single-center reports of multicenter studies Meeting abstracts, editorials, opinion pieces Biomechanical studies, radiographic studies, animal studies, case reports, methodologies

Abbreviations: MAS, minimal access surgery; PLIF, posterior lumbar interbody fusion; TLIF, transforaminal lumbar interbody fusion.
↵a MAS was conducted through a tube using cylindrical soft tissue retractors in 16 studies. In one case, MAS was conducted endoscopically and compared with a conventional open approach.

View popup

Table 2

Characteristics of included studies.

Study	Study Type and Data Source	Country	Mean Age, y		n		Follow-up, mo	Diagnosis	Surgery Type
Study	Study Type and Data Source	Country	MIS	Open	MIS	Open	Follow-up, mo	Diagnosis	Surgery Type
Slotman 199812	RC cost analysis from hospital charges database	USA	37	42	40	31	34	DDD and LS	Open vs lap discectomy
VD Akker 201113	CEA using Euro QOL 5D from randomized controlled trial	Holland	NR	NR	155	159	12	DDD and LS	Open vs MIS discectomy
Lucio 201214	PNRC cost analysis from hospital charges database	USA	64	58	109	101	60	DDD, LS, and DLS	1- and 2-level MIS-TLIF vs open TLIF
Parker 201215	PNRC CEA using Euro QOL 5D	USA	50.8	49.75	9	8	24	DDD, LS, and DLS	Single-level MIS-TLIF vs open TLIF
Wang 201216	RC cost analysis	USA	52	56	52	22	NR	DDD, LS, and DLS	1- and 2-level MIS vs open posterior lumbar interbody fusion
Pelton 201217	PNRC cost analysis	USA	51.6	49.8	33	33	24	DDD, LS, and DLS	Open vs MIS single-level TLIF
Udeh 201218	Cost-utility analysis using Euro QOL 5D and decision model analysis	USA	NR	NR	301	292	24	DDD and LS	Open vs MIS laminectomy
Cahill 201319	RC cost analysis from hospital charges database	USA	45	50	48	33	36	DDD and LS	Open vs MIS discectomy
Parker 201320	RC CEA	USA	59.5	54	27	27	24	DDD and LS	Open vs MIS laminectomy
Singh 201321	PNRC cost analysis from hospital costs database	USA	51.6	49.8	33	33	NR	DDD, LS, and DLS	Open vs MIS single-level TLIF
Parker 201422	CEA using Euro QOL 5D from PNRC	USA	53.5	52.6	50	50	24	DDD, LS, and DLS	Single-level open vs MIS-TLIF
Sulaiman 201423	RC cost analysis from hospital charges database	USA	61	56	57	11	12	DDD, LS, and DLS	Open vs MIS-TLIF
Maillard 201424	RC cost-minimization study (cost-effectiveness measured by hospital cost vs reimbursement)	France	50	59	24	22	12	DDD, LS, and DLS	1- and 2-level open vs MIS-TLIF
R’saud 201525	RC CEA using Euro QOL 5D	Canada	57	55	37	41	24	DDD, LS, and DLS	1- and 2-level open vs MIS-TLIF
Vertuani 201526	CEA using Euro QOL 5D and decision model analysis	UK/Italy	NR	NR	NR	NR	NR	DDD, LS, and DLS	1- and 2-level open vs MIS-TLIF
G’hoke 201627	RC CEA using Euro QOL 5D	USA	57	58	29	45	24	DDD and LS	1-level open vs MIS-TLIF
Djurasovic 201928	PNRC using EQ-5D and SF-6D	USA	57.6	57.0	33	33	12	DLS and DDD	1- and 2-level open vs MIS-TLIF

Abbreviations: CEA, cost-effectiveness analysis; DDD, degenerative disc disease; DLS, degenerative lumbar spondylolisthesis; LS, Lumbar stenosis; MIS, minimally invasive surgery; NR, not recorded/reported; PNRC, prospective nonrandomized cohort study; RC, retrospective cohort study; TLIF, transforaminal lumbar interbody fusion.

View popup

Table 3

Risk of bias assessment.

Study	Clear Definition of Study Population	Clear Definition of Outcomes and Outcome Assessment	Independent Assessment of Outcome Parameters	Random Sequence Generation (Selection Bias)	Allocation Concealment (Selection Bias)	Blinding of participants and Personnel (Performance Bias)	Sufficient Duration of Follow-Up	No Selective Loss During Follow-Up	Important Cofounders and Prognostic Factors Identified
Slotman 199812	+	+	+	×	×	×	+	+	+
Van den Akker 201113	+	+	+	+	+	+	×	+	+
Rampersaud 201125	+	+	+	×	×	×	+	×	+
Wang 201216	+	+	+	×	×	×	+	+	+
Pelton 201217	+	+	+	×	×	×	×	+	+
Parker 201215	+	+	+	×	×	×	+	+	+
Lucio 201214	+	+	+	×	×	×	+	+	+
Udeh 201318	+	+	+	×	×	×	+	+	+
Cahill 201319	+	+	+	×	×	×	+	+	+
Parker 201320	+	+	+	×	×	×	+	+	+
Parker 201422	+	+	+	×	×	×	+	+	+
Singh 201421	+	+	+	×	×	×	+	+	+
Sulaiman 201423	+	+	+	×	×	×	+	×	+
Maillard 201524	+	+	+	×	×	×	×	+	+
Vertuani 201526	+	+	+	×	×	×	+	+	+
Gandhoke 201627	+	+	+	×	×	×	+	+	+
Djurasovic 201928	+	+	+	×	×	×	+	+	+

Note: “+” indicates presence and “x” indicates absence of the given quality.

View popup

Table 4

Study quality assessment—QHES Score.

Study	Study Type and Data Source	Country	Surgery Type	QHES Score	Quality of Economic Study^{^a}
Slotman 199812	RC cost analysis from hospital charges database	USA	Open vs lap discectomy	24	Poor
VD Akker 201113	CEA using Euro QOL 5D from randomized controlled trial	Holland	Open vs MIS discectomy	79	Good
Lucio 201214	PNRC cost analysis from hospital charges database	USA	1- and 2-level MIS-TLIF vs open TLIF	37	Poor
Parker 201215	PNRC CEA using Euro QOL 5D	USA	Single-level MIS-TLIF vs open TLIF	57	Fair
Wang 201216	RC cost analysis	USA	1- and 2-level MIS vs open posterior lumbar interbody fusion	41	Fair
Pelton 201217	PNRC cost analysis	USA	Open vs MIS single-level TLIF	47	Fair
Udeh 201218	Cost-utility analysis using Euro QOL 5D and decision model analysis	USA	Open vs MIS laminectomy	76	Good
Cahill 201319	RC cost analysis from hospital charges database	USA	Open vs MIS discectomy	29	Poor
Parker 201320	RC CEA	USA	Open vs MIS laminectomy	55	Fair
Singh 201321	PNRC cost analysis from hospital costs database	USA	Open vs MIS single-level TLIF	37	Poor
Parker 201422	CEA using Euro QOL 5D from PNRC	USA	Single-level open vs MIS-TLIF	75	Good
Sulaiman 201423	RC cost analysis from hospital charges database	USA	Open vs MIS-TLIF	26	Poor
Maillard 201424	RC cost-minimization study (cost-effectiveness measured by hospital cost vs reimbursement)	France	Open vs MIS-TLIF (max 3 levels)	62	Fair
R’saud 201525	RC CEA using Euro QOL 5D	Canada	1- and 2-level open vs MIS-TLIF	74	Good
Vertuani 201526	CEA using Euro QOL 5D and decision model analysis	UK/Italy	1- and 2-level open vs MIS-TLIF	74	Good
G’hoke 201627	RC CEA using Euro QOL 5D	USA	1-level open vs MIS-TLIF	68	Fair
Djurasovic 201928	PNRC using EQ-5D and SF-6D	USA	1- and 2-level Open vs MIS-TLIF	71	Fair

Abbreviations: CEA, cost-effectiveness analysis; MIS, minimally invasive surgery; PNRC, prospective nonrandomized cohort study; QHES, Quality of Health Economic studies; RC, retrospective cohort study; TLIF, transforaminal lumbar interbody fusion.
Note: Components are weighted by importance (as concluded by expert health economists) to yield a score from 0 to 100 (lowest to highest quality). Literature has suggested that a score of 75–100 points indicates a high-quality economic study. The QHES does not provide insight into study external validity (generalizability) nor does it directly assess the validity of clinical assumptions and inputs.
↵a QHES is a well-validated practical quantitative tool for appraising the quality of cost-effectiveness studies. It assesses multiple aspects of economic study design and reporting to determine internal validity.

View popup

Table 5

Summary of economic evaluation.

Study	Study Type & Data Source	Country	Economic Perspective	Discount Rate	Indirect Costs?	Total Cost	Difference Between MIS and Open Cost (% Difference)^a	MIS Cost per QALY^b	Open Surgery Cost per QALY^b	Change in QALY
Slotman 199812	RC cost analysis from hospital charges database	USA	Payer (user charges)	NR	N	MIS: 5723 (IP) : 4405 (OP)	IP: 1469 (20.4%) OP: 2787 (38.8% P < 0.01)	NR	NR	NR
Slotman 199812	RC cost analysis from hospital charges database	USA	Payer (user charges)	NR	N	Open: 7192	IP: 1469 (20.4%) OP: 2787 (38.8% P < 0.01)	NR	NR	NR
VD Akker 201113	CEA using Euro QOL 5D from randomized controlled trial	Holland	Societal (direct + indirect)	0%	Y	MIS: €33,706	€2972 (19.6%)	42,665	38,321	MIS: 0.790
VD Akker 201113	CEA using Euro QOL 5D from randomized controlled trial	Holland	Societal (direct + indirect)	0%	Y	Open: €30,734	€2972 (19.6%)	42,665	38,321	Open: 0.802 (P = 0.47)
Lucio 201214	PNRC cost analysis from hospital charges database	USA	Payer (user charged)	N/A	N	MIS: 24,320	2735 (10.1%) (P = 0.029)	NR	NR	NR
Lucio 201214	PNRC cost analysis from hospital charges database	USA	Payer (user charged)	N/A	N	Open: 27,055	2735 (10.1%) (P = 0.029)	NR	NR	NR
Parker 201215	PNRC CEA using Euro QOL 5D	USA	Payer (user charges)	N/A	Y	MIS: 35,996	8731 (19.5%) (P = 0.18)	71,992	109,090	MIS: 0.5
Parker 201215	PNRC CEA using Euro QOL 5D	USA	Payer (user charges)	N/A	Y	Open: 44,727	8731 (19.5%) (P = 0.18)	71,992	109,090	Open: 0.41 (P = 0.17)
Wang 201216	RC cost analysis	USA	Payer (Medicare/user charges)	N/A	N	MIS: 70,159 (SL) 87,454 (TL)	SL: 8285 (10.6%) (P = 0.027)	NR	NR	NR
Wang 201216	RC cost analysis	USA	Payer (Medicare/user charges)	N/A	N	Open: 78,444 (SL) 108,843 (TL)	TL: 21,389 (19.7%) (P = 0.071)	NR	NR	NR
Pelton 201217	PNRC cost analysis	USA	Payer (user charges)	N/A	Y	MIS: 28,060 (WC) 29,429 (NWC)	WC: 5602 (17.1%) (P = 0.031)	NR	NR	NR
Pelton 201217	PNRC cost analysis	USA	Payer (user charges)	N/A	Y	Open: 33,862 (WC) 32,998 (NWC)	NWC: 3569 (10.8%)(P < 0.001)	NR	NR	NR
Udeh 201218	Cost-utility analysis using Euro QOL 5D and decision model analysis	USA	Payer (Medicare/user charges)	3%	N	MIS: 5458	8313 (60.3%)	43,760	125,985	MIS: 0.131
Udeh 201218		USA	Payer (Medicare/user charges)	3%	N	Open: 13,771	8313 (60.3%)	43,760	125,985	Open: 0.17
Cahill 201319	RC cost analysis from hospital charges database	USA	Payer (user charges)	N/A	N	MIS: 22,358	5576 (19.6%) (P = 0.01)	NR	NR	NR
Cahill 201319	RC cost analysis from hospital charges database	USA	Payer (user charges)	N/A	N	Open: 27,811	5576 (19.6%) (P = 0.01)	NR	NR	NR
Parker 201320	RC CEA	USA	Payer (Medicare/user charges)	N/A	Y	MIS: 23,109	2311 (9.0%) P = 0.21	32095.8	35305.8	MIS: 0.36
Parker 201320	RC CEA	USA	Payer (Medicare/user charges)	N/A	Y	Open: 25,420	2311 (9.0%) P = 0.21	32095.8	35305.8	Open: 0.36
Singh 201321	PNRC cost analysis from hospital costs database	USA	Hospital costs	N/A	N	MIS: 19,512	4038 (17.1%) (P < 0.001)	NR	NR	NR
Singh 201321	PNRC cost analysis from hospital costs database	USA	Hospital costs	N/A	N	Open: 23,550	4038 (17.1%) (P < 0.001)	NR	NR	NR
Parker 201422	CEA using Euro QOL 5D from PNRC	USA	Payer (Medicare/user charges)	N/A	Y	MIS: 27,621	821 (2.9%) P = 0.5	35,825	40,924	MIS: 0.77
Parker 201422	CEA using Euro QOL 5D from PNRC	USA	Payer (Medicare/user charges)	N/A	Y	Open: 28,442	821 (2.9%) P = 0.5	35,825	40,924	Open: 0.69 (P > 0.05)
Sulaiman 201423	RC cost analysis from hospital charges database	USA	Hospital costs	N/A	N	MIS: 19,078	18,603 (49.4%)	NR	NR	NR
Sulaiman 201423	RC cost analysis from hospital charges database	USA	Hospital costs	N/A	N	Open: 37,681	18,603 (49.4%)	NR	NR	NR
Maillard 201424	RC cost-minimization study (cost-effectiveness measured by hospital cost vs reimbursement)	France	Hospital costs	N/A	N	MIS: €7893	2180 (21.6%)	1139 gain per patient^c	620 loss per patient^a ( P = 0.021)	NR
Maillard 201424		France	Hospital costs	N/A	N	Open: €10,073	2180 (21.6%)	1139 gain per patient^c	620 loss per patient^a ( P = 0.021)	NR
R’saud 201525	RC CEA using Euro QOL 5D	Canada	Hospital costs	5%	N	MIS: 14,183	4450 (23.9%) P = 0.0009	70,915 (2 y)	122,585 (2 y)	MIS: 0.113
R’saud 201525	RC CEA using Euro QOL 5D	Canada	Hospital costs	5%	N	Open: 18,633	4450 (23.9%) P = 0.0009	70,915 (2 y)	122,585 (2 y)	Open: 0.079 (P = 0.08)
Vertuani 201526	CEA using Euro QOL 5D & decision model analysis	UK/Italy^d	Payer (user charged)	N/A	N	MIS: €13,399 (UK) €10,012 (Italy)	UK: 1666 (11.4%)	UK MIS €18,609	UK open: €22,154	NR
Vertuani 201526	CEA using Euro QOL 5D & decision model analysis	UK/Italy^d	Payer (user charged)	N/A	N	Open: €15,065 (UK): €10,985 (Italy)	Italy: 973 (8.9%)	Italy MIS: €13,903	Italy open: €16154.41	NR
G’hoke 201627	RC CEA using Euro QOL 5D	USA	Societal	N/A	Y	MIS: 45,574	1506 (3.3%) (P = 0.96)	75956.67	65773.13	MIS: 0.60
G’hoke 201627	RC CEA using Euro QOL 5D	USA	Societal	N/A	Y	Open: 44,068	1506 (3.3%) (P = 0.96)	75956.67	65773.13	Open: 0.67 (P = 0.3)
Djurasovic 201928	PNRC using EQ-5D and SF-6D from prospective hospital database	USA	Hospital costs	n/a	N	MIS: 15,867	1745 (9.9%)	101,711	124907.80	MIS: 0.156
Djurasovic 201928		USA	Hospital costs	n/a	N	Open: 17,612	1745 (9.9%)	101,711	124907.80	Open: 0.141

Abbreviations: CEA, cost-effectiveness analysis; IP, inpatient; MIS, minimally invasive surgery; NR, not reported/recorded discounted rate; NWC, no workers compensation; OP, outpatient (home on day of surgery); PNRC, prospective nonrandomized cohort study; QALY , quality adjusted life year; RC, retrospective cohort study; SL, single level; TL, two level; WC, workers compensation.
↵a Standard denotes that MIS total cost is less than that of open procedure. Bold font denotes that MIS total cost is more than that of open procedure.
↵b In USD unless otherwise stated.
↵c Maillard et al measured cost-effectiveness as the difference between cost and hospital reimbursement.
↵d UK costs were taken from the UK NHS reference cost list (2015). Costs for Italy and the unit cost estimates for surgical equipment and consumables were obtained from a microcosting study performed in 2 Italian hospitals (ref).