• Siobhan Reilly, Claudia Miranda-Castillo, Reem Malouf, Juanita Hoe, Sandeep Toot, David Challis, and Martin Orrell.
• Division of Health Research, Faculty of Health and Medicine, Lancaster University, C07 Furness Building, Lancaster, UK, LA1 4YG.
• Cochrane Db Syst Rev. 2015 Jan 5; 1 (1): CD008345CD008345.

BackgroundOver 35 million people are estimated to be living with dementia in the world and the societal costs are very high. Case management is a widely used and strongly promoted complex intervention for organising and co-ordinating care at the level of the individual, with the aim of providing long-term care for people with dementia in the community as an alternative to early admission to a care home or hospital.ObjectivesTo evaluate the effectiveness of case management approaches to home support for people with dementia, from the perspective of the different people involved (patients, carers, and staff) compared with other forms of treatment, including 'treatment as usual', standard community treatment and other non-case management interventions.Search MethodsWe searched the following databases up to 31 December 2013: ALOIS, the Specialised Register of the Cochrane Dementia and Cognitive Improvement Group,The Cochrane Library, MEDLINE, EMBASE, PsycINFO, CINAHL, LILACS, Web of Science (including Science Citation Index Expanded (SCI-EXPANDED) and Social Science Citation Index), Campbell Collaboration/SORO database and the Specialised Register of the Cochrane Effective Practice and Organisation of Care Group. We updated this search in March 2014 but results have not yet been incorporated.Selection CriteriaWe include randomised controlled trials (RCTs) of case management interventions for people with dementia living in the community and their carers. We screened interventions to ensure that they focused on planning and co-ordination of care.Data Collection And AnalysisWe used standard methodological procedures as required by The Cochrane Collaboration. Two review authors independently extracted data and made 'Risk of bias' assessments using Cochrane criteria. For continuous outcomes, we used the mean difference (MD) or standardised mean difference (SMD) between groups along with its confidence interval (95% CI). We applied a fixed- or random-effects model as appropriate. For binary or dichotomous data, we generated the corresponding odds ratio (OR) with 95% CI. We assessed heterogeneity by the I² statistic.Main ResultsWe include 13 RCTs involving 9615 participants with dementia in the review. Case management interventions in studies varied. We found low to moderate overall risk of bias; 69% of studies were at high risk for performance bias.The case management group were significantly less likely to be institutionalised (admissions to residential or nursing homes) at six months (OR 0.82, 95% CI 0.69 to 0.98, n = 5741, 6 RCTs, I² = 0%, P = 0.02) and at 18 months (OR 0.25, 95% CI 0.10 to 0.61, n = 363, 4 RCTs, I² = 0%, P = 0.003). However, the effects at 10 - 12 months (OR 0.95, 95% CI 0.83 to 1.08, n = 5990, 9 RCTs, I² = 48%, P = 0.39) and 24 months (OR 1.03, 95% CI 0.52 to 2.03, n = 201, 2 RCTs, I² = 0%, P = 0.94) were uncertain. There was evidence from one trial of a reduction in the number of days per month in a residential home or hospital unit in the case management group at six months (MD -5.80, 95% CI -7.93 to -3.67, n = 88, 1 RCT, P < 0.0001) and at 12 months (MD -7.70, 95% CI -9.38 to -6.02, n = 88, 1 RCT, P < 0.0001). One trial reported the length of time until participants were institutionalised at 12 months and the effects were uncertain (hazard ratio (HR): 0.66, 95% CI 0.38 to 1.14, P = 0.14). There was no difference in the number of people admitted to hospital at six (4 RCTs, 439 participants), 12 (5 RCTs, 585 participants) and 18 months (5 RCTs, 613 participants). For mortality at 4 - 6, 12, 18 - 24 and 36 months, and for participants' or carers' quality of life at 4, 6, 12 and 18 months, there were no significant effects. There was some evidence of benefits in carer burden at six months (SMD -0.07, 95% CI -0.12 to -0.01, n = 4601, 4 RCTs, I² = 26%, P = 0.03) but the effects at 12 or 18 months were uncertain. Additionally, some evidence indicated case management was more effective at reducing behaviour disturbance at 18 months (SMD -0.35, 95% CI -0.63 to -0.07, n = 206, 2 RCTs I² = 0%, P = 0.01) but effects were uncertain at four (2 RCTs), six (4 RCTs) or 12 months (5 RCTs).The case management group showed a small significant improvement in carer depression at 18 months (SMD -0.08, 95% CI -0.16 to -0.01, n = 2888, 3 RCTs, I² = 0%, P = 0.03). Conversely, the case management group showed greater improvement in carer well-being in a single study at six months (MD -2.20 CI CI -4.14 to -0.26, n = 65, 1 RCT, P = 0.03) but the effects were uncertain at 12 or 18 months. There was some evidence that case management reduced the total cost of services at 12 months (SMD -0.07, 95% CI -0.12 to -0.02, n = 5276, 2 RCTs, P = 0.01) and incurred lower dollar expenditure for the total three years (MD= -705.00, 95% CI -1170.31 to -239.69, n = 5170, 1 RCT, P = 0.003). Data on a number of outcomes consistently indicated that the intervention group received significantly more community services.Authors' ConclusionsThere is some evidence that case management is beneficial at improving some outcomes at certain time points, both in the person with dementia and in their carer. However, there was considerable heterogeneity between the interventions, outcomes measured and time points across the 13 included RCTs. There was some evidence from good-quality studies to suggest that admissions to care homes and overall healthcare costs are reduced in the medium term; however, the results at longer points of follow-up were uncertain. There was not enough evidence to clearly assess whether case management could delay institutionalisation in care homes. There were uncertain results in patient depression, functional abilities and cognition. Further work should be undertaken to investigate what components of case management are associated with improvement in outcomes. Increased consistency in measures of outcome would support future meta-analysis.

23 keywords...

hide…