OT_152: Pierse, White & Riggs (2019) — Healthy Homes Initiative Outcomes Evaluation Service: Initial Analysis of…

OT_152: Pierse, White & Riggs (2019) — Healthy Homes Initiative Outcomes Evaluation Service: Initial Analysis of…

Source

https://www.health.govt.nz/publications/healthy-homes-initiative-outcomes-evaluation-service-initial-analysis-of-health-outcomes-interim — original source (opens in a new tab; the file is not redistributed)

Pierse, White & Riggs (2019) — Healthy Homes Initiative Outcomes Evaluation Service: Initial Analysis of Health Outcomes

Pierse N, White M & Riggs L. 2019. Healthy Homes Initiative Outcomes Evaluation Service: Initial analysis of health outcomes. Interim Report, 7 August 2019. He Kāinga Oranga (Housing & Health Research Programme, University of Otago Wellington) with Motu Economic and Public Policy Research; report to the Ministry of Health (co-funded by the Ministry of Health, Housing New Zealand and the Ministry of Housing and Urban Development). HDEC ethics ref 15/STH/138/AM02.

Interim, one-year, referred-child-only — a conservative floor

This is the Phase-1 interim analysis: a before-after (pre-post) study of a single year post-intervention, for the referred child only, with no control population (deferred to the Phase-2 IDI evaluation). The authors state that restricting the analysis to the major health effects for one child per household under-estimates the programme’s true effect. The headline figures should be read as a conservative lower bound, not a central estimate.

Summary

The upstream primary behind LIT_044’s Healthy Homes Initiative (HHI) claims. It is an interim (Phase-1) before-after evaluation of the HHI — a cross-agency programme delivering warm/dry-housing interventions (insulation, curtains, heating, minor repairs, mould kits, housing relocations, and referrals) to low-income whānau with children aged 0–5 or pregnant women, across 11 District Health Boards. Using an evaluation sample of 1,608 referrals (from 4,093 supplied), the study links each referred child’s National Health Index to the National Minimum Dataset (hospital discharges) and the Pharmaceutical Collection, and compares health events in the year before vs the year after the intervention, adjusting for age and (for hospitalisations) selection bias. Extrapolated to all 15,330 referrals received by December 2018, the HHI is estimated to have prevented 1,533 hospitalisations, 9,443 GP visits and 8,784 pharmaceutical dispensings in the first year post-intervention, averting ~29.5m** in public health-care costs over three years against a programme cost of **~19.2m ($1,205/family), giving a return on investment in Year 2. For Neobiome Intelligence the value is a NZ-authoritative, government-commissioned avoided-public-cost benchmark for the health economics of housing-quality investment (D04/I01/I07) — the primary to cite at calc time in place of the LIT_044 review.

Key claims

- claim: "The Healthy Homes Initiative (HHI) was established between December 2013 and March 2015 and covers 11 District Health Boards (Auckland, Waitematā, Counties Manukau, Northland, Waikato, Hutt Valley, Capital & Coast, Lakes, Bay of Plenty, Hawke's Bay and Tairāwhiti). As at 30 December 2018 the programme had received 15,330 eligible referrals and delivered over 40,000 interventions to low-income households."
  source_location: "Background to the HHIs (p.7); Conclusion (p.6, p.20)"
- claim: "The evaluation sample was 1,608 referrals, selected from 4,093 referrals with a primary client that met initial eligibility. The sample was young (40.8% aged 2–5 at earliest intervention) and disproportionately Māori (55.2%) and Pacific (36.6%); by tenure 47.5% were in Housing New Zealand homes, 37.9% in private market rentals and 10% owner-occupied. Earliest intervention date January 2014, latest December 2017 (median October 2016)."
  source_location: "Executive Summary (p.4); Table 1 Characteristics of evaluation sample population (p.13)"
- claim: "After adjusting for age and selection bias, the HHI is estimated to have prevented 160.78 hospitalisations in the sample of 1,608 (rate 0.100 per referral, 95% CI 0.00498–0.184), extrapolating to 1,533 prevented hospitalisations (95% CI 76–2,820) across all 15,330 referrals in the first year post-intervention. Post-intervention hospitalisations were on average 0.69 nights shorter (CI 0.346–1.191) and $541 less costly (CI 9.84–1,073) than in the year prior."
  source_location: "Executive Summary (p.5); Table 4 all-cause hospitalisations (p.15–16); Results: Costs and Benefits (p.17)"
- claim: "After adjusting for age, the HHI is estimated to have prevented 990.17 GP visits in the sample (rate 0.616 per referral, CI 0.563–0.668) → 9,443.28 in the wider population (CI 8,630–10,240), and 921.17 pharmaceutical dispensings in the sample (rate 0.573 per referral, CI 0.354–0.792) → 8,784.09 in the population (CI 5,426–12,141). The age effect on pharmaceutical dispensings was unusually large."
  source_location: "Executive Summary (p.5); Table 5 GP Visits (p.16); Table 6 Pharmaceuticals dispensed (p.16–17)"
- claim: "Programme costs were estimated at $1,205 per family (staffing/delivery, from the 2018 Funding Model Review), and for the 15,330 families served by the programme through December 2018 the total programme cost to the Ministry of Health is estimated at $19,173,581 (rounded to ~$18.5 million in the ROI statement). These costs exclude the cost of providing interventions (e.g. heaters, insulation), which are more often met through donations of funds/time or other programmes."
  source_location: "Executive Summary (p.5); Programme cost data (p.9); Results: Costs and Benefits (p.17)"
- claim: "Total health-care costs averted by the HHI are estimated at $29,473,374 over the three years post-intervention (Year 1 $10,402,150; Years 2–3 discounted at Treasury's 6% rate), summarised as averting approximately $30 million over three years and $10.4 million in the first year. The programme's return on investment is expected to be realised in Year 2 (recouped in less than two years). Unit avoided costs used: hospitalisation $4,090, reduced-severity hospitalisation $541, GP visit $80 (Treasury CBAx Tool), pharmaceutical dispensing $8.45."
  source_location: "Table 7 Health Care Costs Averted by the Healthy Homes Initiative (p.18); Executive Summary (p.6)"
- claim: "Across all referrals the average referred child had an estimated 0.1 fewer hospitalisations, 0.6 fewer GP visits and 0.6 fewer pharmaceuticals dispensed in the 12 months post-intervention; the reduction in nights spent in hospital attributable to the programme is approximately 7,510 (from both prevented hospitalisations and reduced severity). The design is a before-after (pre-post) comparison of the year either side of the intervention, adjusted for an age effect (all outcomes) and, for hospitalisations only, a selection bias (implied selection bias 228.96 hospitalisations arising from prior-hospitalisation eligibility criteria); uncertainty estimated by bootstrap."
  source_location: "Discussion (p.18); Methodology — Corrective adjustments / Estimating uncertainty (p.12–13); Tables 2–4 (p.14–16)"
- claim: "As a benchmark, EECA's Warm Up New Zealand insulation scheme delivered a reduction of 0.06 hospitalisations per child in a sample of low-income families with children (the most cost-effective subgroup, benefit-cost 15:1); the HHI's reduction of 0.10 hospitalisations per referral in this interim analysis is described as a very complementary finding."
  source_location: "Discussion — Comparison with other housing intervention studies (p.19)"

Neobiome Intelligence relevance

• Avoided-public-cost benchmark for housing-quality investment (I01, D04): the ~29.5m three-year averted health-care cost against a ~19.2m programme cost (ROI in Year 2) is a NZ-authoritative, government-commissioned figure for the health economics of warm/dry housing — the avoided-cost half of the case for high-performance habitat in a community costing. Cite this primary (not the LIT_044 review) when the HHI numbers land in an NI calculation. OT_152
• Health-outcome magnitude of housing interventions (I07, D04): 1,533 prevented hospitalisations, 9,443 GP visits and 8,784 dispensings across 15,330 referrals in the first year quantifies how much of the poor-housing hospital burden warm/dry housing provision actually removes — the basic-needs stake behind the warmth/dryness target. OT_152
• Sibling to the insulation-BCR primary (D04/I01): together with OT_087 (Warmer Kiwi Homes 4.66:1 insulation BCR) this is one of the two upstream primaries behind LIT_044’s housing-health economics — OT_087 is the single-measure insulation case, this is the whole-programme, multi-intervention case. OT_152
• Conservatism caveat for the model: interim, one-year, referred-child-only, no control population — the authors state this under-estimates the true effect. Treat the figures as a conservative floor, and note that intervention hardware costs (heaters/insulation) are excluded from the $1,205/family programme cost. OT_152

Research targets

Documents to retrieve

• RT_208 RESOLVED by this source. This is the Pierse et al. (2019) HHI evaluation that RT_208 sought — the upstream primary behind LIT_044’s HHI figures. All three headline numbers confirmed (1,533 prevented hospitalisations; 10,402,150 Year-1 and 29,473,374 three-year costs averted). Residual: the per-region breakdown RT_208 also hoped for does not exist in this interim report (the analysis is aggregate across the 1,608 sample); the per-region / all-household breakdown is promised in the Phase-2 IDI evaluation, which is not yet published — noted here rather than opened as a new RT (per-region cost modelling is a nice-to-have, and national figures are already in hand).

Research gaps

• None new. (The report’s own next-phase agenda — control population, longer follow-up, all household members, per-intervention effects via the IDI — is an HHI-evaluation roadmap, not an NI model gap.)

Notes

• Source-internal cost arithmetic inconsistency (minor, flagged not fixed). The report states a per-family cost of 1,205** and a total programme cost of **19,173,581 for 15,330 families (p.5, p.17), but 1,205 × 15,330 = 18,472,650, not 19,173,581 (and 19,173,581 ÷ 15,330 ≈ 1,251/family). The Executive Summary separately notes only **10,326** families "had been assessed and completed a housing intervention plan" (p.5), which reconciles with neither. The ROI comparison rounds the programme cost to "**approximately 18.5 million**” (closer to the 1,205 × 15,330 product). All figures are reported here verbatim as the source states them; the ~0.7m discrepancy does not affect the Year-2 ROI conclusion (costs averted ~29.5m over 3 yr ≫ ~18.5–19.2m cost).
• Cross-reference: LIT_044’s HHI key_claim (claim #6, “prevented ~1533 hospitalisations, 9443 GP visits and 8784 filled prescriptions … ~10.4m … ~30m over 3 years”) is fully confirmed against this primary — no correction to LIT_044 needed; the LIT_044 “Documents to retrieve” Pierse entry is updated to RESOLVED → OT_152.
• The interim report contains no figures beyond p.21; pages 1–21 constitute the full document (title, acknowledgements, contents, executive summary, methods, results, discussion, conclusion). No appendices were included in the retrieved PDF (Appendix E robustness checks are referenced on p.18 but not present in this file).

Connections

Links to

Sources (2): LIT_044 · OT_087

Referenced by

SSI indicators (2): I01: Financial & Economic Self-Sufficiency · I07: Fulfilment of Basic Needs

Sources (1): LIT_044

EDT domains (1): D04: Sustainable Habitat & Building Technology