Fig 1.
Examples of popular methods to ‘integrate’ care [3] within the health system [5].
Table 1.
Outcome measures.
Table 2.
Subgroup analyses.
Fig 2.
PRISMA flow diagram of study selection [56].
Table 3.
Demographics of included studies.
N/R = Not Reported; N/A = Not Applicable.
Table 4.
Context of included studies.
Table 5.
Details of interventions.
Table 6.
Quality of included studies.
Fig 3.
Forrest plot for self-assessed health status outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Fig 4.
Forrest plot for mortality outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Fig 5.
Forrest plot for total cost of services outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Fig 6.
Forrest plot for utilisation of primary and non-specialist care outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Fig 7.
Forrest plot for utilisation of secondary care outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Fig 8.
Forrest plot for patient satisfaction outcome.
Effect estimates are the standardised mean difference, where the solid vertical line at 0 indicates no effect. Effect estimates are based on a random-effects model. Each subtotal shows the overall effect estimate for the time-period indicated.
Table 7.
No significant differences between subgroups (p<0.05). Note: Positive effect size favours case management for all measures.