Dataset Information

Using Transportability to Understand Differences in Mediation Mechanisms Across Trial Sites of a Housing Voucher Experiment.

ABSTRACT:

Background

Randomized trials may have different effects in different settings. Moving to Opportunity (MTO), a housing experiment, is one such example. Previously, we examined the extent to which MTO's overall effects on adolescent substance use and mental health outcomes were transportable across the sites to disentangle the contributions of differences in population composition versus differences in contextual factors to site differences. However, to further understand reasons for different site effects, it may be beneficial to examine mediation mechanisms and the degree to which they too are transportable across sites.

Methods

We used longitudinal data from MTO youth. We examined mediators summarizing aspects of the school environment over the 10-15 year follow-up. Outcomes of past-year substance use, mental health, and risk behavior were assessed at the final timepoint when participants were 10-20 years old. We used doubly robust and efficient substitution estimators to estimate (1) indirect effects by MTO site and (2) transported indirect effects from one site to another.

Results

Differences in indirect effect estimates were most pronounced between Chicago and Los Angeles. Using transport estimators to account for differences in baseline covariates, likelihood of using the voucher to move, and mediator distributions partially to fully accounted for site differences in indirect effect estimates in 10 of the 12 pathways examined.

Conclusions

Using transport estimators can provide an evidence-based approach for understanding the extent to which differences in compositional factors contribute to differences in indirect effect estimates across sites, and ultimately, to understanding why interventions may have different effects when applied to new populations.

SUBMITTER: Rudolph KE

PROVIDER: S-EPMC7269870 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:AimsTo test how a housing voucher generating residential mobility to lower-poverty neighborhoods, compared with public housing controls, influenced adolescent binge drinking, and whether gender modified effects.DesignA multi-site household-level three-arm randomized trial of a housing intervention executed 1994-98, evaluated 2001-02.SettingFive US cities: Baltimore, MD; Boston, MA; Chicago, IL; Los Angeles, CA; and New York, NY.ParticipantsA total of 3537 adolescents in 4248 low-income eligible families were randomized; 2829 adolescents were analyzed at the interim evaluation (1950 in treatment; 879 in the control group). Attrition bias was accounted for with a 3-in-10 oversampling of hard-to-reach participants (effective response rate: 89%).InterventionsThe Moving to Opportunity (MTO) trial randomized volunteer low-income families in public housing to receive (1) rental subsidies redeemable in neighborhoods with < 10% tract poverty plus housing counseling, (2) unrestricted Section 8 rental subsidies or (3) to remain in public housing. We pooled the subsidy ('treatment') groups because they were conceptually similar and there was no evidence of statistical differences between groups on binge drinking.MeasurementsPrimary outcome: past month binge drinking (five or more drinks in one sitting).FindingsAdolescent binge drinking prevalence was 3.9% for treatment and 3.2% for control. The intention-to-treat (ITT) main effect of subsidy treatment (versus control) on binge drinking was non-significant, but treatment effects were different for girls and boys (treatment-gender interaction P = 0.002). MTO treatment reduced girls' binge drinking [odds ratio (OR) = 0.48, 95% confidence interval (CI) = 0.24-0.96, P = 0.037], but increased boys' binge drinking (OR = 2.37, 95% CI = 1.13-4.97, P = 0.023), compared with controls. Results were similar for secondary alcohol outcomes. Instrumental variable (IV) results adjusting for treatment compliance were comparable with ITT, but larger.ConclusionsA housing subsidy treatment that enables low-income families to move from public to private housing appears to lessen girls' binge drinking but increases boys' binge drinking, compared with controls.

Project description:ImportanceAlthough neighborhoods are thought to be an important health determinant, evidence for the relationship between neighborhood poverty and health care use is limited, as prior studies have largely used observational data without an experimental design.ObjectiveTo examine whether housing policies that reduce exposure to high-poverty neighborhoods were associated with differences in long-term hospital use among adults and children.Design, setting, and participantsExploratory analysis of the Moving to Opportunity for Fair Housing Demonstration Program, a randomized social experiment conducted in 5 US cities. From 1994 to 1998, 4604 families in public housing were randomized to 1 of 3 groups: a control condition, a traditional Section 8 voucher toward rental costs in the private market, or a voucher that could only be used in low-poverty neighborhoods. Participants were linked to all-payer hospital discharge data (1995 through 2014 or 2015) and Medicaid data (1999 through 2009). The final follow-up date ranged from 11 to 21 years after randomization.ExposuresReceipt of a traditional or low-poverty voucher vs control group.Main outcomes and measuresRates of hospitalizations and hospital days, and hospital spending.ResultsAmong 4602 eligible individuals randomized as adults, 4072 (88.5%) were linked to health data (mean age, 33 years [SD, 9.0 years]; 98% female; median follow-up, 11 years). There were no significant differences in primary outcomes among adults randomized to receive a voucher compared with the control group (unadjusted hospitalization rate, 14.0 vs 14.7 per 100 person-years, adjusted incidence rate ratio [IRR], 0.95 [95% CI, 0.84-1.08; P?=?.45]; hospital days, 62.8 vs 67.0 per 100 person-years; IRR, 0.93 [95% CI, 0.77-1.13; P?=?.46]; yearly spending, $2075 vs $1977; adjusted difference, -$129 [95% CI, -$497 to $239; P?=?.49]). Among 11?290 eligible individuals randomized as children, 9118 (80.8%) were linked to health data (mean age, 8 years [SD, 4.6 years]; 49% female; median follow-up, 11 years). Receipt of a housing voucher during childhood was significantly associated with lower hospitalization rates (6.3 vs 7.3 per 100 person-years; IRR, 0.85 [95% CI, 0.73-0.99; P?=?.03]) and yearly inpatient spending ($633 vs $785; adjusted difference, -$143 [95% CI, -$256 to -$31; P?=?.01]) and no significant difference in hospital days (25.7 vs 28.8 per 100 person-years; IRR, 0.92 [95% CI, 0.77-1.11; P?=?.41]).Conclusions and relevanceIn this exploratory analysis of a randomized housing voucher intervention, adults who received a housing voucher did not experience significant differences in hospital use or spending. Receipt of a voucher during childhood was significantly associated with lower rates of hospitalization and less inpatient spending during long-term follow-up.