Network meta-analyses performed by contracting companies and commissioned by industry

Table 1 Results of the 19 contracting companies that indicated that they perform network meta-analysis

	Company	Start date of first NMA	NMAs completed	NMAs published	NMA conducted/year	NMA published/year	Ratio published/conducted
1	Mapi Group (incl. Health Technology Analysts and Optum)	Nov-05	88	33	9.5	3.6	0.38
2	Abacus International/Decision Research Group	Jun-07	54	10	7.0	1.3	0.19
3	Redwood Outcomes (now part of Precision Health Economics)	Sept-13	51	5	36	3.5	0.10
4	Evidera/UBC^c	Nov-08	50	3	8.0	0.48	0.06
5	Kleijnen Systematic Reviews Ltd^c	Jan-08	20	14	2.8	2.0	0.70
6	Amaris	Jan-12	15	1	4.9	0.32	0.07
7	Symmetron Ltd	Nov-08	13	2	2.1	0.32	0.15
8	RTI Health Solutions	Jun-00	12	13	0.82	0.89	1.08
9	Medignition Inc.	Aug-08	6	6	0.92	0.92	1.00
10	Augmentium Pharma Consulting	Jan-05	5	4	0.50	0.40	0.80
11	AHEAD	Jun-10	2	2	0.43	0.43	1.00
12	Xintera Consulting	Jan-09	2	2	0.33	0.33	1.00
13	CEMKA-EVAL	May-12	2	0	0.73	0.0	0.0
14	McMDC Ltd	Sep-09	2	1	0.37	0.18	0.50
15	Pharmacoeconomics and Outcomes Research Iberia	Jan-10	2	1	0.39	0.20	0.50
16	BeSyRe Bekkering Systematic Reviews	May-10	1	1	0.21	0.21	1.00
17	David Hoaglin (consulting statistician)	Jan-09	1	1	0.16	0.16	1.00
18	Company 1^c	NP	NP	NP	NP	NP	0.04
19	Company 2^c	NP	NP	NP^a	NP	NP	0.08
	Total	–	476^b	106^b	6.6	1.5	0.22

Companies are sorted based on the number of network meta-analyses (NMA) conducted
NP not presented due to confidentiality
^aSince no list of references was available from Company 2, we used the number of papers identified by our literature search
^bMeta-analyses conducted or published by more than one contracting company are double-counted, but these are uncommon. The sum of 106 for published network meta-analyses corresponds to 102 papers, because four meta-analyses were co-authored by two contracting companies. The exact level of redundancy for conducted meta-analyses is not correct, but is likely to be similarly low
^cThese companies gave an approximation for the number of conducted network meta-analyses. For calculations the approximation was considered as the actual number of network meta-analyses

ISSN: 2046-4053