“…Most of the studies (44 studies or 66.7%) were secondary data analyses of existing databases, [ 42 , 44 , 45 , 47 , 50 , 51 , 54 , 57 , 59 – 63 , 65 , 66 , 68 , 69 , 71 , 72 , 74 – 76 , 80 , 82 , 83 , 87 , 90 , 93 – 96 , 100 – 102 ]; of which, ten of these studies used econometric analytic techniques [ 41 , 43 , 56 , 64 , 77 , 84 , 86 , 91 , 97 , 98 ]. Other study designs included nine cross-sectional surveys [ 40 , 48 , 55 , 58 , 79 , 81 , 85 , 103 , 105 ], five longitudinal cohort studies [ 46 , 49 , 70 , 78 , 88 ], four retrospective cohort studies [ 73 , 89 , 99 , 104 ], and one each of a case control [ 53 ], quasi experimental [ 52 ], cluster randomized controlled trial [ 92 ] and quality improvement study [ 67 ]. Fifty-six studies (84.9%) controlled for health status in some way, either in study design or using statistical techniques.…”