The Future of Children, Princeton - Brookings: Providing research and analysis to promote effective policies and programs for children.

The process assessment focuses on family engagement and retention, service delivery, and quality of care as described in HSP training manuals and in state contracts with HSP sites. As noted previously, the model calls for three to five years of home visiting of decreasing intensity as families achieve milestones in healthy functioning. New families are expected to need weekly visits for most, if not all, of the first year. Home visitors aim to develop individualized service plans with each family at least every six months. Home visit content is individualized to help the family achieve six-month goals, such as graduating from high school and getting infant immunizations. Child developmental screenings and mother-child interaction assessments are to be carried out at four months and periodically thereafter. Quality of care is defined in two ways: (1) maternal satisfaction with the home visitor, and (2) home visitor success in identifying and responding to problems that require intervention.

Despite early and frequent attempts to contact referred families, home visitors found it challenging to do so. For 29% of referred families, the home visitors first tried to contact the family within a day of receiving the referral; for 78% of families, the first contact attempt was made within a week. Even so, home visitors never directly contacted 3% of mothers and never visited another 9%. For the 88% with at least one visit, the time from birth to the first visit was 23 days on average.

It was also challenging to keep families in the program. A program site considered a family active as long as staff were visiting the family or felt it was reasonable to continue to try to establish or maintain family involvement. Attrition rates were higher than anticipated, with 10% of referred families considered inactive by their program sites by the time the child was three months old, 30% by six months, 44% by nine months, and 51% by 12 months.

Attrition is a common problem in family support programs.⁹ It is possible that Healthy Start attrition rates may be somewhat higher than rates in other programs because HSP uses universal screening to identify at-risk families and outreach to bring them into home visiting. For programs in which universal screening is used, higher attrition can probably be expected than in programs in which families request service.

Understanding the reasons for attrition is important for developing better ways to identify, engage, and retain families in service. Reasons for families leaving Healthy Start prematurely are listed in Table 2. Family refusal was by far the most common reason for attrition. Most families refusing service did so early on, either before the first visit or before developing an individual service plan. Still, 11% of families refused service after completing such a plan.

Fully 9% of families dropped out of the program because of problems of accessibility—6% moved to parts of the state outside of the HSP catchment areas or moved to catchment areas where intake was closed, and 3% had work or school schedules that conflicted with the home visitors' work hours. Attrition might be reduced if HSP service availability were expanded.

Few families were visited weekly, as expected in the HSP model. Table 3 illustrates that for all referred families, there were an average of 13 home visits in the infant's first year. Considering only the time from the home visiting referral to either program discharge or the child's first birthday, 29% of families were visited at least every two weeks and 51% at least every three weeks. Families still active at one year had an average of 22 visits, with nearly half visited at least every two weeks.

Home visitors developed individual service plans with 71% of families, screened 55% of the index children, and assessed parent-child interactions in 47% of all referred families. Among families active at one year, 97% had individual service plans, 92% had infant developmental screenings, and 84% had assessments of parent-child interactions.

A trusting relationship is key in helping vulnerable families accept supportive and educational services. Mothers rated home visitors using a self-administered questionnaire completed at the conclusion of the annual data collection. The overall rating was derived from the mother's level of agreement with 25 statements describing the home visitor's empathy, understanding, respect for the family, trustworthiness, and ability to motivate and guide the family toward independence and effective parenting.¹⁰

At the one-year follow-up, 76% of the mothers in the HSP group recalled having had a home visitor. Recall was related to the frequency of visits and current status; 96% of those considered active by their program sites at one year recalled having had a home visitor, as compared with 65% of inactive families with 12 or more visits and 47% of inactive families with fewer than 12 visits.

At one year, 28% of control group mothers also reported having had a home visitor, although none of these families had HSP home visitors. Other programs in Hawaii provide home-based services, and data collection in the ongoing third-year follow-up will investigate which programs provided such services to control group families.

In the interim, study results presented in Figure 2 show that mothers in the HSP group rated their home visitors much more favorably than did control group mothers. Thus, although HSP home visitors found it difficult to engage and retain families, they did convey to most mothers their sensitivity to family issues and their commitment to helping them be good parents.

As illustrated in Figure 3, the three agencies taking part in the evaluation varied greatly in family retention rates, home visit frequency, and home visitor ratings. HFSC had the highest family retention rate, the lowest refusal rate, and the highest number of visits among all families. PACT had the highest number of visits among families still active at one year. CFS had the highest home visitor ratings, although all three agencies had high ratings.

HSP network members believe that the agency differences in refusal rates and visit frequency reflect differences in philosophy. All three agencies aim to provide the same level of service, but PACT views the entire family, more than the index child, as its primary client. Thus, its home visitors are likely to honor a family's change of heart about whether to accept home visiting, and they focus their efforts on providing services to more receptive families.

The other two agencies expect that many at-risk families will be reluctant to engage in home visiting, but they believe this underscores the need for aggressive outreach. They regard the engagement of an isolated family as more important than honoring a family's inclination to be left alone. Thus, home visitors in these two agencies are encouraged to continue to engage families by telephone and in person in instances in which home visitors at PACT would be advised to send a letter offering a final chance to accept services.

Why CFS home visitors are rated most highly is harder to explain. HSP network agencies have suggested that this difference might relate to agency differences in the content of home visits, supervision, and the matching of home visitors and families. Subsequent reports will address this issue more thoroughly and determine which aspects of the process are the most important determinants of visitor ratings.

Program administrators can guide outreach and training efforts better if they understand how program engagement varies among family subgroups. Multivariable logistic regression was used to examine HSP's success in engaging families in at least 12 home visits in the first year. The results indicated that families were more likely to have had at least 12 visits if the father was violent, substance abusing, and at extremely high risk; if the mother had substance use problems; if the mother did not use violence unilaterally as a means of dealing with conflict with her partner; and if the mother was not at extremely high risk. These findings suggest that it is not simply "difficult" families that drop out or lower-risk families that self-select out of service. The three administering agencies did not differ significantly in the types of families retained.