TABLE 1.
Characteristics of included studies
| Authors (year) | Study design | N | Women (%) | Age (range and/or mean [SD]) | Social factors under study | Was the relationship for social factors significant? | Summary of the relationship found between social factors and outcomes | Length of follow-up (months) |
|---|---|---|---|---|---|---|---|---|
| Cobey et al., 1976 | Prospective observational cohort | 89 | 75 | 65–91 | Participation—getting out of the home prior to injury | Yes | Participation prior to fracture was correlated with recovery as measured by modification of a functional activity scale developed by Katz | 6 |
| Magaziner et al., 1990 | Prospective observational cohort | 340 | 83.50 | 78.1 ± 7.1 | Social network | Yes | Greater contact with ones social network is positively associated with recovery in three areas (walking ability, physical dependence and instrumental dependence). | 12 |
| Egan et al., 1992 | Prospective observational cohort | 61 | 78.70 | 76.6 ± 7.4 | Role loss (investigated using the Role Checklist) and social support (investigated using part 2 of the Personal Resource Questionnaire) | No | Dependence in ADL’s at home was not significantly related to role loss (different roles in life). | 0.7 (3 weeks) |
| Marottoli et al., 1992 | Prospective observational cohort | 118 | 72.03 | 65+; 78.2 | Social support (social network size, number of sources of emotional or task support, marital status and social activities) | No | At 6 weeks and 6-months fewer sources of emotional support at baseline was not statistically significant as a predictor of better physical function. | 6 |
| Marottoli et al., 1994 | Prospective observational cohort | 120 | 71.67 | 65+; not given | Social network and support measures (social network size, number of sources of emotional or task support, marital statusand social activities) | No for the majority of factors | Measures were not associated with an increase or decrease risk of mortality. Being unmarried was the only social support measure that trended towards institutionalisation. | 6 |
| Oh & Feldt, 2000 | Prospective observational cohort | 70 | 88.60 | 65+; 84.12 | Social supportas measured using a modified Norbeck Social Support Questionnaire | Yes | Perceptions of network size and instrumental support at discharge were correlated with functional status, as measured by the Functional Status Index, at 2 months following discharge. | 2 |
| Cree et al., 2001 | Prospective observational cohort | 367 | 79 | 65+; 82 | Social support determined by whether the respondent had someone to rely on for help when needed; measured by the Older Americans Resources and Services (OARS) | Yes | Functional dependence among patients of low mental function is predicted by social support. | 3 |
| Cresci, 2001 | Cross-sectional exploratory |
73 | 100.0 | 65+; 81 ±6.87 | Informational support (used the modified Inventory of Socially Supportive Behaviors (ISSB) provided by natural support systems | Yes | Informational support was a significant predictor of post injury functional status | N.A. |
| Allegrante et al., 2007 | Randomised control trial | 59 | 76.27 | 65+; 77 ±8 | Multi-faceted rehab approach: a motivational video and social contact with an age-matched peer who had recovered from hip fracture | Yes | Intervention arm had a significant positive change in the role-physical scale as compared to the control. | 6 |
| Mortimore et al., 2008 | Prospective observational cohort | 674 | 77.40 | 65+; 81.1 ± 7.4 | Interaction with social network | Yes | Infrequent contact with family or friends before hip fracture are at higher risk of dying than are those who have frequent contact. | 24 |
| Morghen et al., 2011 | Prospective observational cohort | 280 | 88.40 | 65+; 80.2 ± 6.8 | Living arrangement | No | Living alone was not significant predictor of failure to recover walking independence at discharge | 12 |
| Sylliaas et al., 2012 | Prospective observational cohort | 277 | 82 | 65+; 82.4 ± 6.5 | Living arrangement | No | Living arrangement was not a significant predictor of either ADL or IADL | 3 |
| Gambatesa et al., 2013 | Randomised control Trial | 40 | 92.50 | 70+; 80.8 ±6.5 | Counselling (as a social support) | Yes | Counselling had a positive impact on health-related quality of life on all patients, but in a more relevant way if patients were low functioning upon admittance to the ward. | 1 |
| Orive et al., 2016 | Prospective observational cohort | 740 | 79.07 | 65 and older; not given | Home status, institutional support at baseline, independence level and income | Yes | Predictors of worsening pain at 6 months and/or 18 months included living in a home-care situation or nursing home before the fracture. Social predictors of deterioration in function at 6 months and/or 8 months included lower income. | 18 |
| Hongisto et al., 2016 | Prospective observational cohort | 841 | 78.10 | 65 and older; 81.9 ± 6.77 | Living with somebody, previous living arrangement | Yes | Living with somebody prior to hip fracture was protective against institutionalisation at 1 year. Institutionalisation at 1 and 4 months after hip fracture considerably increased the risk of death and permanent institutionalisation 12 months after hip fracture. | 12 |
| Kristensen etal.,2017 | Prospective population-based cohort study | 25,354 | 49 | 65 and older; not given | Socioeconomic markers: education, income, cohabiting status and migrant status | Yes (for some) | Patients with higher education had a lower 30-day mortality risk. Level of family income was associated with lower 30-day mortality. Patients with both high education and high income had a lower risk of acute readmission. | 1 |
| Landeiro et al., 2016 | Prospective observational cohort | 278 | 79.50 | 75+; 85.5 ±5.8 | Social isolation | Yes | Being isolated or at a high risk of social isolation, was significantly associated with delayed discharges. | 0.4 (13 days on average) |
| Shin et al., 2016 | Retrospective cohort study | 5,441 | 71 | 65 and older; not given | SES (income) | Yes | Mean survival time was longer as the income level increased | 120 (this is retrospective database study with 11 years of follow-up for some patients) |
| Thorne et al., 2016 (Wales) | Record linkage study —database study | 11,098 | 73.25 | 80.4 ± 11.1 | Social demographics: income, employment, health and disability, education, barriers to housing and services, crime, living environment | Yes for social deprivation and mortality rates | Social deprivation (based on the social demographics indicated) was significantly associated with increased mortality in the most deprived quintile compared with the least deprived quin- tile at 90 and 365 days in Wales | 12 |
| Thorne et al., 2016 (England) | Record linkage study—database study | 171,570 | 72.85 | 80.7 ± 11.6 | Social demographics: income, employment, health and disability, education, barriers to housing and services, crime, living environment | Yes for social deprivation and mortality rates | Social deprivation was significantly associated with increased mortality in the most deprived quintile compared with the least deprived quintile at 30, 90 and 365 days in England. | 12 |
Abbreviations: ADL, activities of daily living; IADL, instrumental activities of daily living.