Table 3.

Measures of linkage and impact of studies on linkage outcomes

Study	Main linkage-related outcome	Direct or indirect measure of linkage to primary care	Measurement method	Impact on linkage?
Braun 2015 [36]	Change in cancer screening behaviours	Indirect	Baseline and exit surveys	Yes. Significant increase in screening. 57.0% had had a Papanicolaou test, 61.7% had had a mammogram, 54.4% had had a prostate-specific antigen test and 43% had had a flexible sigmoidoscopy or colonoscopy at the study exit. Specific number linked to primary care as a direct result of program not known
Dennis 2015 [17]	Access to mainstream health services	Indirect	Feedback from service users via semi-structured interviews, postal surveys and community forums	Yes. Feedback from surveys and qualitative comments suggested that brokerage service successfully linked Aboriginal Australians with primary care providers who were able to meet their needs. Specific number linked to primary care as a direct result of program not known
Han 2009 [37]	Change in breast screening behaviour	Indirect	Baseline and follow up questionnaires	Yes. Significant increase in screening. Screening increases at 6 month follow up were 31.9% for mammography, 23% for clinical breast examination and 36.2% for breast self-examination compared with baseline. Intervention involved referring participants to health providers but specific numbers seen by providers not given.
Hiatt 2008 [38]	Change in cancer screening behaviours	Indirect	Baseline and follow up surveys for random community samples (baseline and follow up groups were not the same)	No. No significant effects on screening behaviour between intervention and control groups. Specific linkage to primary care not measured.
Findley 2011 [32]	Number of participants obtaining an asthma action plan from a health professional during the intervention period	Direct	Baseline and follow up questionnaires	Yes. 76% of participants obtained an asthma action plan from a health professional at Los Angeles and New York site and 100% obtained an asthma action plan at Philadelphia site
Johnson 2012 [39]	Changes in number of office visits to primary care providers and specialists before, during and after the intervention period	Indirect	Utilisation data extracted from Molina Healthcare of New Mexico (Medicaid Managed Care provider organisation) records	Yes. CHWs assisted members to establish a primary care medical home- office visits to primary care providers and specialists dropped by half in non-CHW group but remained relatively stable in the CHW group. Specific number linked to primary care as a direct result of program not known
Jordan 2013 [33]	Number of participants who were linked to primary care in the community during intervention period	Direct	Transitional care coordinators contact community primary care provider to verify linkages to care and document this in the Electronic Health Record and monthly summary reports prepared with information on number linked to care	Yes. % of linkages to primary care/releases to the community: 70% in 2009, 75% in 2010 and 73% in 2011. The monthly average was 73% linked to primary care (monthly mean = 98).
Krantz 2013 [30]	Number of participants referred to medical or lifestyle resources during intervention period	Indirect	Number of referrals tracked through study database	Unclear. 53.5% of participants at risk for CHD received medical or lifestyle resource referrals. Specific number of these who made appointments with provider/s not known.
Mason 2013 [34]	Number of community members who had filled out a mammography interest/screening form who received a mammogram at the collaborating health facility in intervention period	Direct	Programmatic tracking forms used to keep track of mammogram appointment status	No. Only 21% of women who completed a mammography interest/screening form received mammograms at the collaborating health facility (a comparatively smaller proportion than other screening studies in the review)
Wang 2012 [35]	Number seen in primary care clinic at least once in the 12 month follow-up period and number seen 2 or more times	Direct	Utilisation data at follow-up extracted from the UCSF Clinical and Translational Science Institute’s The Health Records Electronic Data Set	Yes. More than 60% of participants were seen in primary care clinic at least once in the 12-month follow-up period, and 42% were seen 2 or more times. After 12 months of follow up, 37.7% of Transitions Clinic participants and 47.1% of Enhanced Primary Care participants made 2 or more visits to their assigned primary clinic.
Whitley 2011 [31]	Number referred to primary care and/or other resources during intervention period	Indirect	Number of referrals tracked through study database	Unclear. Over one-third of participants were referred to primary care and/or other resources. Specific number of these who made appointments with provider/s not known.