From: Which interactions matter in economic evaluations? A systematic review and simulation study
 | Completed studies | Protocols and terminated studies | Overall | ||
---|---|---|---|---|---|
Publication yeara | 2009 | 18% (7/40) | 50% (7/14) | 26% (14/54) | |
2008 | 10% (4/40) | 14% (2/14) | 11% (6/54) | ||
2007 | 5% (2/40) | 29% (4/14) | 11% (6/54) | ||
2006 | 10% (4/40) | 0% (0/14) | 7% (4/54) | ||
2005 | 15% (6/40) | 0% (0/14) | 11% (6/54) | ||
2000–2004 | 28% (11/40) | 7% (1/14) | 6% (3/54) | ||
Pre-2000 | 15% (6/40) | 0% (0/14) | 9% (5/54) | ||
Mainly/entirely publicly funded | 75% (30/40) | 100% (14/14) | 81% (44/54) | ||
Country where research done | United Kingdom | 28% (11/40) | 43% (6/14) | 31% (17/54) | |
United States | 38% (15/40) | 7% (1/14) | 30% (16/54) | ||
Other individual country | 23% (9/40) | 29% (4/14) | 24% (13/54) | ||
Multinational | 13% (5/40) | 21% (3/14) | 15% (8/54) | ||
Disease area | Cardiovascular disease | 45% (18/40) | 0% (0/14) | 33% (18/54) | |
Musculoskeletal | 13% (5/40) | 14% (2/14) | 13% (7/54) | ||
Cancer | 15% (6/40) | 0% (0/40) | 11% (6/54) | ||
Smoking, drugs or alcohol | 13% (5/40) | 29% (4/14) | 17% (9/54) | ||
Other | 25% (10/40) | 57% (8/14) | 33% (18/54) | ||
Factors target different diseases | 3% (1/40) | 14% (2/14) | 6% (3/54) | ||
Type of interventionb | Pharmaceutical | 45% (18/40) | 36% (5/14) | 43% (23/54) | |
Physiotherapy, exercise and related | 15% (6/40) | 29% (4/14) | 19% (10/54) | ||
Nutrition | 8% (3/40) | 14% (2/14) | 9% (5/54) | ||
Surgery | 13% (5/40) | 7% (1/14) | 11% (6/54) | ||
Training, counselling, incentives and logistical interventions | 50% (20/40) | 43% (6/14) | 48% (26/54) | ||
Sample sizec | Mean (range) | 3420 (126, 20,536) | 1371 (132, 4733) | 3002 (126, 20,536) | |
< 200 patients | 13% (5/39) | 7% (1/14) | 11% (6/53) | ||
200–999 patients | 38% (15/39) | 36% (5/14) | 38% (20/53) | ||
> 1000 patients | 49% (19/39) | 29% (4/14) | 43% (23/53) | ||
Type of economic evaluation | Cost-effectiveness analysis | 68% (27/40) | 0% (0/14) | 50% (27/54) | |
Cost-effectiveness and cost-utility analysis | 18% (7/40) | 43% (6/14) | 24% (13/54) | ||
Cost-utility analysis | 8% (3/40) | 29% (4/14) | 13% (7/54) | ||
Cost-consequence analysis | 8% (3/40) | 0% (0/14) | 6% (3/54) | ||
Cost-utility and cost-benefit analysis | 0% (0/40) | 7% (1/14) | 2% (1/54) | ||
Depends on results | 0% (0/40) | 7% (1/14) | 2% (1/54) | ||
Not stated | 0% (0/40) | 14% (2/14) | 4% (2/54) | ||
Cluster-randomised | 13% (5/40) | 43% (6/14) | 20% (11/54) | ||
Type of design | Full factorial | 88% (35/40) | 86% (12/14) | 87% (47/54) | |
Partial factorial | 10% (4/40) | 7% (1/14) | 9% (5/54) | ||
Incomplete factorial | 3% (1/40) | 7% (1/14) | 4% (2/54) | ||
Factorial matrix size | 2 factors & 2 levels2 (e.g. 2 × 2, 2 × 2 + 1, or double 2 × 2) | 78% (31/40) | 71% (10/14) | 76% (41/54) | |
> 2 factors each with 2 levels (e.g. 2x2x2) | 10% (4/40) | 0% (0/14) | 7% (4/54) | ||
2 factors with > 2 levels on ≥1 factor (e.g. 3 × 2) | 10% (4/40) | 21% (3/14) | 13% (7/54) | ||
> 2 factors with > 2 levels on ≥1 factor (e.g. 3x2x2) | 3% (1/40) | 7% (1/14) | 4% (2/54) | ||
Primary clinical endpoint | Assumed no interaction | 53% (21/40) | 29% (4/14) | 46% (25/54) | |
Included interaction term | 23% (9/40) | 7% (1/14) | 19% (10/54) | ||
Unclear | 25% (10/40) | 64% (9/14) | 35% (19/54) | ||
Base case economic evaluation | Assumed no interaction | 40% (16/40) | 7% (1/14) | 31% (17/54) | |
Included interaction term | 53% (21/40) | 0% (0/14) | 39% (21/54) | ||
Unclear | 8% (3/40) | 93% (13/14) | 30% (16/54) | ||
Presentation of uncertainty | None | 48% (19/40) | – | 48% (19/40) | |
Present using confidence intervals and/or scatter graphs | 15% (6/40) | – | 15% (6/40) | ||
Cost-effectiveness acceptability curves | Pair-wise comparison(s) | 23% (9/40) | – | 23% (9/40) | |
Multiple comparison(s) | 8% (3/40) | – | 8% (3/40) | ||
Pair-wise and multiple comparison(s) | 8% (3/40) | – | 8% (3/40) |