Table 1:
Economic analyses alongside empirical studies of parenting, early years and school-based interventions promoting mental health and well-being
| Bibliographic information | Intervention (I) and comparator (C) | Target population and duration of economic analysis | Study design | Cost results | Mental health-related effectiveness results | Perspective/price year | Synthesis of costs and effectiveness data |
|---|---|---|---|---|---|---|---|
| (Cunningham et al., 1995), Canada | I: Large group community-based parenting programmes | Parents of 150 pre-school/kindergarten children at high risk of developing conduct disorders | RCT | Community-based groups were reported to be more than three times as much as clinic/individual parenting sessions | Community group had a significantly greater number of solutions to problems than control groups (p< 0.05) Significantly better in reducing behavioural problems at home compared with the clinic group (p< 0.05). Community group reported greater improvement than the clinic group, but significantly better parental sense of competence in the clinic control group (p< 0.05) | Health sector and travel costs | No synthesis of costs and benefits. Community-based group reported have better outcomes than clinic-based programmes and to be six times more cost-effective because of higher number of people reached by group sessions |
| C: Clinic-based individual parenting programmes or 6 months waiting list | 6 months | CCA | CAD. Price year not stated | ||||
| (Edwards et al., 2007), Wales | I: The Webster-Stratton Incredible Years group parenting programme | Parents of 116 children aged 36–59 months at risk of developing conduct disorders | Pragmatic | The mean cost per child attending the parenting group: £934 for 8 children and £1289 for 12 children containing initial costs and materials for training group leaders. | Risk of conduct disorder linked with child behaviour. Significant improvement in mean intensity scores for child behaviour on Eyeberg scale in the intervention group of 27 points compared with no change in the control group (p< 0.0001) | A multiagency public sector perspective: health, special educational and social services | Incremental cost per five point improvement on the Eyeberg intensity scale would be £73. Given a ceiling ratio of £100 per point change 83.9% likelihood of being cost-effective |
| C: 6 months waiting list | 6 months | RCT | Incremental costs of all health, social and special education services were £1992.29 compared with £49.14 in the control group | 2004 GBP | Estimated to cost £5486 to bring child with highest intensity score below clinical cut-off for risk of developing conduct disorders | ||
| CEA | |||||||
| (Foster, 2010), USA | I: Fast Track intervention: multi-year, multi-component prevention programme targeting antisocial behaviour and violence. Includes curriculum based on the PATHS programme which focuses on social and emotional learning. Includes parent training, home visiting, academic tutoring, social skills training | 891 children identified at first year of entry to school system and provided intervention services over a 10-year period | RCT | Intervention cost $58 000 per child. Average health service costs (excluding programme costs) per child were $2450 in the intervention group | Focus on broad range of long-term outcomes that are associated with onset of conduct disorder in childhood: delinquency, school failure and use of school services, risk of substance abuse. No significant intervention effects were found | Public purse | No ratio reported the author states that ‘the most intensive psychosocial intervention ever fielded did not produce meaningful and consistent effects on costly outcomes. The lack of effects through high school suggests that the intervention will not become cost-effective as participants progress through adulthood’ (Foster, 2010) |
| C: No intervention | CEA | 2004. USD | |||||
| (Foster et al., 2008), USA | Population wide implementation of multi-level Triple P intervention. (see Mihalopoulos et al., 2007) | Parents and children in nine counties in South Carolina | Ongoing RCT in South Carolina | The costs for universal media and communication components: less than $0.75 per child in population | Outcomes of intervention are not reported here. Instead a threshold analysis conducted to identify costs that could be avoided if programme effective. Thresholds in line with those reported in previous studies | Programme costs plus costs to participants of various events | Estimated that the cost of implementing Triple P could be recovered in 1 year by a 10% reduction in child abuse and neglect |
| COA | Total costs of providing interventions from levels 2–5 $2, 183, 812 or cost per family of $22 or $11.74 per child | USD. Price year not stated | |||||
| (Foster et al., 2007), USA | I: Incredible Years Programme with three components: a child-based training programme (CT), a parent-based training programme (PT) and a teacher-based training programme (TT). | 459 children aged 3–8 not receiving mental health treatments and their parents | Six RCTs | The total cost per child was $1164 with CT, $1579 with PT, $2713 with CT and PT, $1868 with PT and TT, $1454 with CT and TT and $3003 with CT, PT and TT | Parent–child interaction measured using Dyadic Parent–Child Interactive Coding System–Revised (DPICS-R; observer reported). Preschool behaviour measured using Behar Preschool Behavior Questionnaire (PBQ; teacher reported) used | Intervention costs to health and education system, including travel and refreshments and childcare costs | If payers have willingness to pay of $3000 per unit of improved behaviour on PBQ then PT and TT treatment are most cost-effective, while for values lower than $3000 no treatment was the preferred strategy |
| Each component focused on improving children's behaviour through the promotion of socially appropriate interaction skills. | Data taken from six clinical trials | CEA | Parent–child interaction improved significantly for all intervention groups, except CT only. Preschool behaviour improved significantly all treated groups except for the CT, PT and TT group | 2003 USD | If parent–child interaction improvement then if willingness to pay of $2500 per unit of effectiveness, the CT, PT and TT option was the most cost-effective in almost 70% of cases | ||
| C: Comparisons were made between different combinations of the three components plus no intervention | To end of delivery of Incredible Years programme | ||||||
| (Foster and Jones, 2006, 2007), USA | I: Fast Track intervention: multi-year, multi-component prevention programme targeting antisocial behaviour and violence. Includes curriculum-based on the PATHS programme which focuses on social and emotional learning. Includes parent training, home visiting, academic tutoring, social skills training | 891 children identified at first year of entry to school system and provided intervention services over a 10-year period | RCT | The average cost $58 283 per participant | Diagnosis of conduct disorder using the Diagnostic Interview Schedule for Children Self Report of Delinquency instrument for violence | Public purse | Cost per case of conduct disorder averted: $3 481 433 for all population; $752 103 for high-risk individuals |
| CEA | Effectiveness outcomes are not explicitly reported in paper—only the incremental cost-effectiveness ratios | 2004, USD | Cost per act of inter-personal violence prevented $736 010 | ||||
| C: No intervention | 10 years | Intervention not considered cost effective for lower risk groups Would be cost-effective for highest risk groups if societal willingness to pay above $750 000 |
|||||
| (Hiscock et al., 2007), Australia | I: Advice and education from maternal and child health nurses to improve infant sleep and maternal well-being. | 328 mothers reporting infant sleep problems at 7 months | Cluster RCT | The mean cost for intervention: £96.93 versus control family: £116.79. (non-significant difference) | Significant reduction in reported infant sleep problems at 10 months for the intervention group : 56 versus 68% (p = 0.04) and at 12 months 39 versus 55% (p= 0.007). Significant mean difference in risk of post-natal depression for the intervention group—1.4 on Edinburgh Post Natal Depression Scale (p< 0.007); significantly improved mental health scores on SF-12 for intervention—mean difference 3.9 p< 0.001 | Health-care perspective | Ratio not reported as intervention dominant: lower costs, higher benefits |
| 5 months | CCA | (MCH sleep consultations, other health-care services and interventions costs) | |||||
| C: Usual consultations at Maternal and Child Health Centres | GBP. Price year not stated | ||||||
| (McIntosh et al., 2009) and (Barlow et al., 2007), England | I: An intensive home visiting programme | 131 vulnerable families at risk of abuse and neglect | Multicentre RCT | Health service only: intervention £5685 versus control £3324 | Statistically significant improvement in maternal sensitivity and infant co-operativeness components of the CARE Index outcome measure. Maternal sensitivity 9.27 in the intervention group versus 8.20 in the control group (p= 0.04) | Health and societal perspectives | No ratio assessing cost-effectiveness per unit improvement in maternal sensitivity or infant co-operativeness |
| C: Care as usual | 18 months | CCA | Societal costs: intervention £7120 versus £3874 for control | Infant co-operativeness 9.35 versus 7.92 in the control group (p= 0.02) | 2004 GBP | However, cost per child identified as being at risk of neglect would be at least £55 016 | |
| 0.059 rate increase in (non-significant increase in protection of children from abuse and neglect | |||||||
| (Morrell et al., 2000), England | I: Post-natal support from a community midwifery support workers: practical and emotional support, to help women rest and recover after childbirth | 523 new mothers aged 17 plus | RCT | At 6 months, the intervention group had significantly meant higher costs of £180. (equivalent to cots of support worker) | No evidence of significant difference in health status between groups using SF-36 or in post-natal depression using the Edinburgh Post Natal Depression Scale at 6, 6 weeks or 6 months | Health service | No ratio reported as comparator dominant with lower costs and no difference in outcomes |
| C: Standard midwife care, plus up to 10 visits from support workers during first 28 days | 6 weeks and 6 months | CCA | At 6 months these differences persisted with mean cost of £815 in the intervention group versus £639 in the control group | 1996 GBP | |||
| (Morrell et al., 2009), England | I: Health visitor delivered psychological interventions, cognitive behavioural approach (CBA) or person-centred approach (PCA)+ SSRI | 418 women at high risk of post-natal depression | Pragmatic randomized cluster trial | No significant difference in costs at 6 months between intervention and controls: £339 versus £374 | At 6 months 45.6% of women in the intervention group compared with 33.9% of control found to be at risk of post-natal depression with scores >12 on the Edinburgh Post-Natal Depression Scale (p= 0.028) | NHS and social service perspective | No ratio and intervention dominant with similar or lower costs and better outcomes. In sensitivity analysis 90% chance of being cost-effective if threshold between £20 000–30 000 per QALY gained |
| C: Health visitor usual care | 6 months; analysis at 12 months of small sample only | CCA | SF-6 used to generate Quality Adjusted Life Year values. Incremental gain of 0.003 QALYs in the intervention group (0.026 versus 0.023) | 2005 GBP | In a small sample at 12 months intervention also dominant | ||
| CUA | |||||||
| (Niccols, 2008), Canada | I: Eight session parent group ‘Right From the Start’ (RFTS) to enhance skills in reading infant cues and responding sensitively | 76 mothers of infants | RCT | The mean costs per person per session were significantly lower for intervention: RFTS: $44.04 versus home visiting: $91.26 (p< 0.001) | No significant differences in outcomes on infant attachment security (measured by Attachment-Q set AQS) or maternal sensitivity (measured using Maternal Behaviour Q-score) | Health system plus parental travel costs | No incremental cost-effectiveness ratio as lower cost and better outcomes. Average cost per gain in A QS score for intervention was $430.08 compared with $1283.54. In sensitivity analysis for every $100—Return on investment three to eight times greater than for home visiting |
| C: Routine health visiting | 8 months | CEA | CAD. Price year not stated | ||||
| (Olds et al., 1993), USA | I: Home visiting programme, social support for mother until child is age 2 | 400 new mothers. Emphasis on teenage, single and low-income mothers; but also other mothers | RCT | For whole population incremental programme cost $3246 | Health outcomes reported in other papers, including positive effects on child mental health/risk of abuse/maternal mental health | Societal | Net costs of $1582 per mother for whole population. Net savings of $180 per mother in the low-income group |
| C: Screening for developmental problems at 2 years; free transportation to regular prenatal and well-child care local clinics | 48 months | COA | For low-income population incremental programme cost $3133 | 1980. USD | |||
| Societal | Economic analysis focused on long-term costs of government programmes assumed to be influenced by improved maternal and child health | ||||||
| (Petrou et al., 2006), England | I: Health visitor delivered counseling and support for mother–infant relationship | 151 expectant mothers at high risk of post-natal depression | RCT | Mean intervention group costs per mother–infant pair were £2397 versus £2278 in the control group. Non-significant difference of £119.50 | There was a non-statistically significant difference in time spent with post-natal depression (9.57 weeks in the intervention group versus 11.71 weeks in the control group) | Health and social care perspective | Incremental cost per depression free month gained of £43 |
| C: Routine primary care | 18 months | CEA | 2000; GBP | If willingness to pay of £1000 for preventing 1 month of post-natal depression, intervention 71% chance of being cost-effective (71%) with mean net benefit of £384 | |||
| CBA | |||||||
| (Scott et al., 2010), England | PALS study (Primary Age Learning Skills Trial) | 174 children in very deprived areas of London from diverse ethnic backgrounds (76% were from minority groups) | RCT | The programme cost was £1343 per child. Total cost of the programme was £176 000 | Child behaviour problems (measured through observation and Parent Account of Child Symptoms Schedule. Conduct scale of Strengths and Difficulties Questionnaire (SDQ) also completed. Parenting monitored using approach of Conduct Problems Research Programme. No significant differences in outcomes were reported with the exception that the intervention group had greater use of child centred parenting and more use of calm discipline | Study funder plus health service | No ratio provided. Authors stated programme may need to be designed to increase parent uptake and engagement to be cost-effective |
| I: Basic Incredible Years Parenting Programme (12 weeks) plus 6 weeks manualized SPOKES (Supporting Parents on Kids Education in Schools) Literacy programme to help parents interact with children over books they are using l + SPOKES (6 weeks)→Primary Age Learning Skills (PALS) | CCA | GBP price year not stated | |||||
| C: No intervention | |||||||
| (Wiggins et al., 2004, 2005), England | I: Supportive listening home visits by a support health visitor (SHV) or year of support from community groups (CG) providing drop in sessions, home visiting and/or telephone support | 731 culturally diverse new mothers living in deprived inner city London | RCT | There were no significant differences in total costs between those in SHV, CG and control groups after 12 or 18 months although the interventions tend to be more costly: the 18 month mean costs estimated to be £3255, £3231 and £2915, respectively | Maternal depression was measured at 8 weeks and 14 months post-partum using Edinburgh post-natal depression scale (EPDS). General health questionnaire (GHQ12) used at 20 months post-partum | Public sector, voluntary groups and mothers | No ratio reported as no difference in outcomes found |
| C: Standard health visitor services | 12 and 18 months | CUA | 2000 GBP | No net economic cost or benefit of choosing either of the two interventions or standard health visitor services |
RCT, randomized controlled trial; CBA, cost–benefit analysis; CEA, cost-effectiveness analysis; CCA, cost-consequences analysis; CUA, cost–utility analysis; COA, cost-offset analysis.