Patterns of Visit Attendance in the Nurse–Family Partnership Program

Abstract

Objectives. We examined visit attendance patterns in the Memphis trial of the Nurse–Family Partnership and associations between these patterns and family characteristics, outcomes, and treatment–control differences in outcomes.

Methods. We employed repeated measures latent class analysis to identify attendance patterns among the 228 mothers assigned to receive home nurse visits during pregnancy and until the child was aged 2 years, associated background characteristics, outcomes, and treatment–control differences by visit class. Home visits were conducted from June 1990 to March 1994. We collected outcome data from May 1992 to April 1994 and July 2003 to December 2006.

Results. We identified 3 visit attendance patterns. High attenders (48%) had the most visits and good outcomes. Low attenders (33%) had the most education and the best outcomes. Increasing attenders (18%) had the fewest completed visits during pregnancy, the poorest intake characteristics, and the poorest outcomes. Treatment–control group differences varied by class, with high and low attenders having better outcomes on some measures than did their control group counterparts.

Conclusions. Three patterns were associated with distinct groups of mothers with different long-term outcomes. Further examination and use of patterns to classify mothers and prioritize resources may improve efficiency in the Nurse–Family Partnership.

Low visit attendance is a common problem with preventive interventions¹ and occurs across home visitation programs.^2–5 Missed visits are concerning because they are missed opportunities for providers to educate and support families. The positive outcomes intended for these programs are diminished with shorter duration of involvement⁴ and less contact time between the family and the visitor.⁶ The degree to which low visit attendance is a problem, however, is not easily quantified because there may not be a simple dose–response relationship between visit attendance and outcomes.

Missed visits are associated with participant background characteristics and outcomes inconsistently across studies. Some of these inconsistencies may be attributable to differences in goals, target populations, and service providers. Other apparent inconsistencies may result from assumptions of linear or dichotomized relationships,^5,7–10 in which there is some evidence of nonlinear associations. For instance, in the Elmira, New York, and Memphis, Tennessee, trials of the Nurse–Family Partnership (NFP), completing more visits was associated with lower maternal psychological resources in a linear relationship with an index based on mother’s intelligence, mental health, mastery (the extent to which the mother believes she can control her own life outcomes), and self-efficacy¹¹ as well as a quadratic function with mothers with high and those with low psychological resources both receiving more visits than mothers with average psychological resources.¹²

Lower visit attendance has been associated with lower income^8,10 and education^3,8 and worse mother and child health.¹⁰ However, stress, substance abuse, and mental health problems are associated with higher visit attendance.⁷ In one study, more social support was associated with lower visit attendance,⁷ but in another, being married and living with a partner was associated with higher visit attendance, whereas living alone was associated with lower visit attendance.³ The relationship between psychosocial characteristics and visit attendance is complex.

Completed visits require cooperation between the mother and the nurse; the mother must be available at the scheduled time and the nurse must reach out to build trust and must reschedule missed appointments. Differences in visit patterns may reflect differences in families’ needs, mothers’ abilities to participate, and visitors’ interpretations of those needs. Visitors’ responses to these factors may differ depending on service provider characteristics or the personality match between the mother and nurse.¹³ In one program, sites with visitors who delivered the program with more flexibility had higher retention than did other sites,³ and increasing visitor flexibility increased completed home visits.²

Visit attendance can be measured in many ways, including time to attrition, number of visits completed, and total contact time. Examining visit patterns over time instead of these aggregate measures may uncover nonlinear relationships and provide insight into the families’ experiences. Visit patterns can indicate if there are common times when families drop out or frequently miss visits or if there are other common trajectories. Previous work identifying attendance patterns focused on programs with a fixed number of visits,^8,14 but home visiting programs often have a variable number, depending on the timing of enrollment.

Visit attendance patterns at specific times during an intervention with a variable number of sessions have not been studied, to our knowledge, nor have the relationships between such patterns and outcomes. Identification of distinct visit attendance patterns may help predict which families would benefit from retention efforts and may improve our understanding of complex relationships between dose and outcomes.

NFP is an evidence-based home visitation program implemented nationally¹⁵ in which fewer than half of recommended visits were completed in the original trials and in the current national replication of the program.^2,3 NFP starts during pregnancy and recommends visits until the child’s second birthday.¹⁶ Visit frequency varies over time to support relationship development between the nurse and family and to accommodate shifts in maternal and child health and development over this 2.5-year period.¹⁷ Nurses promote improvements in the mothers’ health and parenting behaviors, economic self-sufficiency, and supportive relationships and link mothers with health and community services.¹⁸

Three randomized controlled trials of the NFP model found consistent improvements in a range of outcomes, including children’s home environments, children’s language development, childhood injuries, and the timing of subsequent pregnancies for mothers.^11,19–21 The intervention affected some outcomes for the whole sample and others (including child’s academic achievement) only for children of mothers with low psychological resources (40%–50% of the samples).

The effects of NFP depend on engagement of the families with the program in accordance with the visit patterns achieved in the original NFP trials. Because of different levels of family need and engagement, we examined 3 questions: (1) Were there discernible variations in completed visit patterns over the course of the program? (2) Were those variations associated with risk and outcomes? (3) Were there intervention–control group differences in outcomes for the subgroups defined by visit patterns?

METHODS

We used data from the Memphis trial of the NFP, which enrolled women in 1990 to 1991. Primiparous women at less than 29 weeks of gestation were recruited at an obstetric clinic for the poor. Participants met 2 of 3 criteria: unemployed, unmarried, less than 12 years of education. The resulting sample was primarily Black (92%), young (64% under 19 years), poor (85% at or below poverty based on U.S. Census Bureau definition), and unmarried (98%).¹¹

The control (n = 515) and intervention (n = 228) groups both received transportation to prenatal care, developmental screening, and referral services for children.¹⁸ The intervention group was offered nurse home visits from enrollment until the child was aged 2 years. Home visits were conducted from June 1990 through March 1994. Nurses aimed to complete all recommended visits but were encouraged to make greater attempts at engaging mothers with the greatest need.¹⁷ Interviewers masked to treatment assignment collected outcome data periodically after the child’s birth until the child was aged 12 years. We examined selected maternal and child outcomes from the 24-month (May 1992 to April 1994) and 12-year (July 2003 to December 2006) assessments. The participants who were interviewed at 12 years had characteristics equivalent to those of the original participants.²¹

We excluded participants who experienced miscarriages (n = 8), stillbirths (n = 2), and neonatal losses (n = 1) from this study, because visits were discontinued. Visit data were missing on 5 mothers. The resulting sample size was 212 for visit pattern analyses. Twelve-year outcome data were missing on 33 children, resulting in a sample size of 179 for those analyses. The control group sample size was 469 for 24-month outcomes and 397 for 12-year outcomes.

Measures

We considered all in-person visits completed visits. The program environment was characterized by whether there was racial concordance with the initial nurse and whether the nurse assigned to the family changed during program participation, primarily because of nurse turnover.

We included maternal age as a 3-level categorical variable (< 17 years, 17–18 years, and > 18 years) to allow a nonlinear relationship, following the original trial stratification.¹¹ Mothers younger than 17 years are at greater risk for having children of low birth weight and benefited more from the program in the previous, Elmira trial. Those aged 18 years and younger are at greater risk for a variety of child health and developmental problems, including limited economic self-sufficiency. Mother’s employment reflected her age, family and economic situation, and availability to attend visits.

Other demographic measures were race, marital status, living alone, attending school, the grade of school completed, contact between the mother and father of the baby (“father involved”), receipt of government assistance, household discretionary income (difference between self-reported income and subsistence standards on the basis of household size), and neighborhood poverty (percentage of households at or below the federal poverty level in the census block).

Substance use included any use of alcohol, tobacco, and illicit drugs; we did not include smoking separately because of low prevalence (< 10%). We created mother’s psychological resources from 4 baseline measures: IQ,²² mental health,²³ mastery,²⁴ and self-efficacy related to behavioral objectives of NFP following Bandura’s model.²⁵ We captured health beliefs through timing of prenatal care initiation, timing of program enrollment, and child-rearing attitudes (higher scores on the Adolescent–Adult Parenting Index indicate higher risk of child maltreatment²⁶).

We selected 5 outcomes affected by the intervention in the Memphis trial and other NFP trials.^{11,20,21,27–31} We measured 2 outcomes near the child’s second birthday: (1) score on the Home Observation for Measurement of the Environment,³² and (2) mother’s subsequent pregnancy. We collected 3 child outcomes near age 12 years. We measured internalizing disorders with the Youth Self-Report³³ and defined them as “overcontrol of emotions—includ[ing] social withdrawal, demand for attention, feelings of worthlessness or inferiority, and dependency.”^34(p677) Following the original study, we dichotomized internalizing disorders to indicate clinical or borderline disorder.²¹ Reading and math achievement were measured using the Peabody Individual Achievement Tests.³⁵ We limited our achievement score analyses to children of mothers with low psychological resources, because the intervention affected only this group.^19,36

Statistical Analysis

We employed repeated measures latent class analysis. We used PROC LATENT CLASS ANALYSIS in SAS version 9.3 (SAS Institute, Cary, NC)³⁷ to identify visit attendance classes (groups of mothers) on the basis of nurse visit patterns. Latent class analysis is used to identify homogeneous subsamples of respondents from a population with a mixture distribution.^38,39 Repeated measures latent class analysis is a nonparametric approach grounded on categorical indicators of class membership that can model patterns of change over time.³⁸ We chose this approach because methods with continuous measures require assumptions of a functional form of patterns (such as polynomial), which were not likely to model visit patterns well.

We created periods so that the recommended frequency of visits was constant within each period (Figure 1). Because there was substantial variation in visit attendance during the first 6 postnatal weeks, we separated this time into 2 periods (1–3 weeks and 4–6 weeks). We calculated the percentage of visits completed compared with recommended for each period and created a dichotomous variable indicating whether the mother attended more than 50% of the recommended visits. We determined the number of classes from model fit statistics (Akaike information criterion⁴⁰ and adjusted Bayesian information criterion⁴¹) as well as interpretability of patterns.

FIGURE 1—

Visit completion trajectories of each class by periods: Nurse–Family Partnership; Memphis, TN; June 1990–March 1994.

Note. The midpoint of each period represents that period on the graph. Period 1 = 1–4 weeks following enrollment; recommended visit frequency = weekly. Period 2 = 5 weeks after enrollment until birth; recommended visit frequency = every 2 weeks. Period 3 = birth through 3 weeks; recommended visit frequency = weekly. Period 4 = 4–6 weeks; recommended visit frequency = weekly. Period 5 = 7 weeks–12 months; recommended visit frequency = every 2 weeks. Period 6 = 12–20 months; recommended visit frequency = every 2 weeks. Period 7 = 21–24 months; recommended visit frequency = monthly.

Because race is associated with attrition,³ we examined models including only Black mothers (90%), tested for measurement invariance by race, and found no substantial differences in visit patterns. To maximize power, we have reported results that include all races.

We used bivariate analysis to identify variables associated with class membership. False discovery rate control is used in multiple hypothesis testing to correct for multiple comparisons. Because of the large number of maternal characteristics included in the bivariate analyses, to control the expected proportion of incorrectly rejected null hypotheses (“false discoveries”), we used the false discovery rate method to determine the appropriate cutoff for statistical significance, for an overall false-positive rate of 5%.⁴² Multiple regression analyses included predictors identified from the bivariate analysis, with P < .10. In cases of significantly correlated variables (P < .05), we selected the variable with the lowest P value.

We examined dichotomous outcomes (subsequent pregnancies, internalizing disorders) using logistic regression and continuous outcomes (Home Observation for Measurement of the Environment, math and reading achievement) using linear regression. Following the original study, we controlled for neighborhood poverty, mother’s child-rearing attitudes, and psychological resources.²¹

We compared outcomes for each intervention group class to the full control group and to matched comparison groups constructed within the control group for each class. We generated class assignment in the control group by predicting each control mother’s likely class on the basis of her intake characteristics, assuming mothers with similar intake characteristics would follow similar visit attendance patterns if they had been given the opportunity. We used multiple imputation with chained equations⁴³ using 100 imputations⁴⁴ and included class membership as a missing variable for each control group mother.

Note that class membership is missing completely at random, because trial randomization determined treatment group assignment. A supplement to this article at http://www.ajph.org contains the variables we used for the imputation model, along with comparisons between the intervention and control groups on these variables. For each class, we compared the intervention group outcomes with the corresponding matched control group outcomes.

Most intake variables had few missing values (≤ 2%). Higher rates of missingness were present for father involved (6%), parenting attitudes (11%), neighborhood poverty (11%), and racial concordance with the nurse (12%). The 12-year outcomes were missing for 16% to 17%.

The Home Observation for Measurement of the Environment was missing for 6%, and subsequent pregnancies were missing for 3%. We did not use imputation for missing data because the overall rate of missing was low and imputation is not appropriate for outcome variables.

On the basis of our findings, we conducted several analyses to further understand visit attendance classes. We examined birth outcomes, the mother’s school and employment after the child’s birth, and the nurse–mother relationship. We used a Nurse–Client Relationship Inventory, which independent interviewers conducted near the child’s second birthday and which consisted of 27 items with a 5-level Likert-scale describing the mother’s relationship with her nurse (Cronbach α = 0.96). Lower values correspond to better relationships.

RESULTS

Both 3- and 4-class models fit the data. We chose the 3-class model on the basis of the small difference between the 3- and 4-class model for both Akaike information criterion and adjusted Bayesian information criterion (Table 1) and the simpler interpretation of fewer classes. Figure 1 illustrates these classes. “High attenders” (48%) attended at least 50% of recommended visits throughout the program. “Increasing attenders” (18%) increased their average visit attendance from 43% just after birth to 96% in the second year. “Low attenders” (33%) attended fewer than 50% of recommended visits in most periods. High attenders had more visits throughout the program (50.1 visits of 61.5 recommended) than did increasing attenders (42.4; P < .003) and low attenders (22.9; P < .001).

TABLE 1—

Model Fit Statistics for Repeated Measures Latent Class Analysis: Nurse–Family Partnership; Memphis, TN; June 1990–March 1994

No. of Classes	AIC	Adjusted BIC	Entropy
1	261.43	262.75	1.00
2	151.85	154.67	0.70
3	138.75	143.07	0.70
4	138.22	144.05	0.71

Note. AIC = Akaike information criterion; BIC = Bayesian information criterion.

The only statistically significant bivariate association between class membership and predictors was highest grade completed in school (Table 2). Low attenders completed the most grades in school, and increasing attenders completed the fewest. In multivariable models (Table 3), high and increasing attendance were both associated with less education than was low attendance.

TABLE 2—

Bivariate Associations Between Class Membership and Mother’s Intake Characteristics: Nurse–Family Partnership; Memphis, TN; June 1990–March 1994

Variable	Class 1: High Attenders,a %, Mean (SD), or Mean (Range)	Class 2: Increasing Attenders,a %, Mean (SD), or Mean (Range)	Class 3: Low Attenders,a %, Mean (SD), or Mean (Range)	Difference Between Classes, Pb
Program
Racial concordance with first nurse	36.3	41.0	42.9	.26
Nurse changed during program	52.5 (50.1)	71.1	54.9	.02
Demographics
Aged < 17 y	43.1 (13–16)	38.5 (13–16)	19.7 (15–16)	.005
Aged > 18 y	34.3 (19–29)	33.3 (19–33)	43.7 (19–26)	.78
Mother’s race is White	8.8	7.7	14.1	.45
Married	2.0	0.0	1.4	.57
Lives alone	1.0	0.0	2.9	.33
In school	64.7	53.8	46.5	.08
Highest completed grade in school	9.9 (2.1)	9.4 (1.9)	10.9 (1.7)	≤ .001*
Employed at enrollment	2.9	12.8	8.5	.13
Father involvedc	84.4	91.7	91.0	.1
Received government assistanced	83.1	83.3	79.3	.42
Household discretionary income	−543.0 (6718.0)	−397.0 (6398.0)	1270.0 (6889.0)	.14
Neighborhood povertye	35.8 (20.5)	40.2 (21.2)	33.7 (18.5)	.05
Substance use during pregnancyf	10.8	15.4	14.1	.78
Psychological characteristics
Mental healthg	99.7 (10.6)	97.5 (8.7)	99.6 (11.7)	.18
Masteryh	98.2 (10.2)	99.7 (10)	101.5 (9.5)	.43
Personal psychological resourcesi	98.7 (11.3)	97 (9.9)	102.2 (10.2)	.02
Health beliefs
Prenatal care, time of first health care visit in wk after last menstrual period	15.4 (6.3)	15.6 (5.2)	17.4 (6.6)	.06
Parenting attitudesj	102.2 (8.8)	101.3 (8.5)	99.4 (8.2)	.06
Estimated gestation at enrollment, wk	16.5 (5.9)	16.5 (5.5)	17.5 (5.5)	.46

Note. Age reference = 17–18 years.

^aHigh attenders attended at least 50% of recommended visits throughout the program. Increasing attenders had low visit attendance early in the program but increased attendance during the first year of the child’s life. Low attenders had high visit attendance before birth but decreased attendance after birth.

^bLog-likelihood test for significance.

^cMother had contact with the child’s father.

^dReceived assistance from government programs, including public assistance, food stamps, Medicaid, and Social Security.

^ePercentage of households at or below federal poverty level (based on the U.S. Census Bureau definition).

^fSelf-reported substance use, including drugs, alcohol, and tobacco.

^gRAND Mental Health Inventory.

^hPearlin Mastery Scale. Mastery is the extent to which the mother believes she can control her own life outcomes.

ⁱPersonal psychological resources on the basis of intelligence (Shipley), mastery, mental health, and efficacy; dichotomized by sample mean for categorical variable.

^jAdult-Adolescent Parenting Inventory. Higher scores are associated with a higher likelihood of child abuse and neglect.

*P < .05, after false discovery rate correction (cutoff = 0.0024).

TABLE 3—

Multinomial Logistic Model Predicting Class Membership From Intake Characteristics: Nurse–Family Partnership; Memphis, TN; June 1990–March 1994

Characteristic	Odds of Being a High vs Increasing Attender,a OR (95% CI)	Odds of Being an Increasing vs Low Attender, OR (95% CI)	Odds of Being a High vs Low Attender, OR (95% CI)
Highest grade completed	1.33 (0.91, 1.94)	0.59* (0.41, 0.84)	0.78* (0.63, 0.97)
Nurse changed during program	0.33 (0.07, 1.46)	2.30 (0.55, 9.60)	0.76 (0.34, 1.72)
First prenatal visit, wk after last menstrual period	0.98 (0.98, 1.01)	0.99 (0.98, 1.01)	0.99 (0.98, 1.002)
Father involvedb	0.13 (0.004, 3.91)	2.64 (0.07, 97.40)	0.35 (0.093, 1.30)

Note. CI = confidence interval; OR = odds ratio. The final model contains variables with uncorrected P < .05 in bivariate analyses. We eliminated age, personal psychological resources, school, parenting attitudes, and neighborhood poverty because of significant correlation with highest grade completed.

^aHigh attenders attended at least 50% of recommended visits throughout the program. Increasing attenders had low visit attendance early in the program but increased attendance during the first year of the child’s life. Low attenders had high visit attendance before birth but decreased attendance after birth.

^bMother had contact with the child’s father.

*P < .05.

For the intervention group, we compared outcomes for each class with each other and detected no significant differences (Table 4; data available as a supplement to this article at http://www.ajph.org). High and low attenders had better home environments than did the full control group (Table 4). High attenders had fewer subsequent pregnancies than did the full control group. We did not see any differences between any of the classes and the full control group for school achievement or internalizing disorder outcomes.

TABLE 4—

Associations Between Class Membership and Outcomes: Nurse–Family Partnership; Memphis, TN; June 1990–April 1994, July 2003–December 2006

Outcome (When Data Collected)	High Attenders,a % or Mean (SE)	Increasing Attenders,a % or Mean (SE)	Low Attenders,a % or Mean (SE)	Full Control Group, % or Mean (SE)
Subsequent pregnancy (24 mo)
Intervention group	41*b	41	28**c
Matched control group subsampled	47	48	47	48
Home environmente (24 mo)
Intervention group	31.7*b,c (5.7)	31.2 (6.8)	32.8**b (5.2)
Matched control group subsampled	30.7 (5.7)	30.2 (5.9)	31.9 (5.5)	31.0 (5.7)
Math achievementf (12 y)
Intervention group	87.3*c (11.7)	83.6 (9.1)	89.9 (9.6)
Matched control group subsampled	86.0 (9.9)	85.9 (10.3)	88.4 (10.5)	86.8 (10.3)
Reading achievementf (12 y)
Intervention group	88.4 (11.6)	86.4 (10.5)	91.6 (11.8)
Matched control group subsampled	89.4 (12.2)	88.7 (12.7)	91.1 (11.2)	89.8 (12.1)
Internalizing problem behaviorsg (12 y)
Intervention group	26	20	27
Matched control group subsampled	31	31	31	32

Note. Intervention group (n = 212) had 48% in high attendance, 18% in increasing attendance, and 33% in low attendance. The control group (n = 514) had imputed class membership of 39% in high attendance, 27% in increasing attendance, and 34% in low attendance. We tested outcomes between classes for the intervention group only and did not find any significant associations.

^aHigh attenders attended at least 50% of recommended visits throughout the program. Increasing attenders had low visit attendance early in the program but increased attendance during the first year of the child’s life. Low attenders had high attendance before birth but decreased attendance after birth.

^bContrast of each class to the full control group.

^cContrast of each class to the matched control group.

^dWe used multiple imputation with chained equations to create matched subsamples of the control group on the basis of intake variables.

^eHome Observation for Measurement of the Environment. Higher scores indicate a home environment that better supports child development.

^fPeabody Individual Achievement Test at 12 years, low-resource only. (We defined low-resource mothers as the bottom 50% of the sample for an aggregate measure of baseline IQ, mastery [the extent to which the mother believes she can control her own life outcomes, and parenting self-efficacy.) Higher scores indicate greater achievement. Intervention group (n = 113) had 52% in high attendance, 20% in increasing attendance, and 27% in low attendance. The average control group per imputation (n = 256) had imputed class membership of 42% in high attendance, 32% in increasing attendance, and 27% in low attendance.

^gYouth Self-Report (Achenbach), percentage of children with scores above the borderline clinical threshold.

*P < .05; **P < .01.

Evidence of successful control group matching is presented in a supplement to this article at http://www.ajph.org. The only differences between the intervention and matched control groups on intake characteristics were that increasing attenders had lower education than did their matched controls and high attenders talked to their baby’s father less often than did their matched controls. High attenders had better home environments and children with higher math achievement than did their matched control group. Low attenders had fewer subsequent pregnancies. No differences were seen between any of the classes and matched control groups for the child’s reading achievement or internalizing disorders.

Increasing attenders had greater likelihood of preterm birth (18%) than did high attenders (4%; P < .005) but not low attenders (8%; P < .24). Babies of increasing attenders had longer hospital stays (11 days; P < .001) than did those of high attenders (3.7) and low attenders (3.7). Increasing attenders had infants of lower gestational age (38.0 weeks; P < .002) than did high (39.7) and low (39.4) attenders. Birth weight (P < .051) and Apgar scores (P < .31) were not associated with class membership (means shown in a supplement to this article at http://www.ajph.org).

Among increasing attenders, the infants of matched controls had lower 5-minute Apgar scores and younger gestational age at birth than did intervention babies (data available as a supplement to this article at http://www.ajph.org). Among low attenders, the infants of matched controls also had younger gestational age than did intervention babies.

School attendance at 12 and 24 months, employment at 12 and 24 months, and subsequent pregnancy by 12 months were not associated with class membership (data available as a supplement to the online version of this article at http://www.ajph.org). High and increasing attenders reported better relationships with their nurses (1.74 and 1.73, respectively) than did low attenders (2.04; P < .002).

DISCUSSION

The 3 visit attendance patterns were associated with predictors and outcomes. High attenders constituted the majority of the mothers in this study. Their intake characteristics and outcomes fell between the other 2 groups. This group most closely followed the recommended visit schedule, and therefore we have used it as the reference group in our discussion.

Low attenders had the highest education. Although no other characteristics reached statistical significance after false discovery rate adjustment, from an exploratory perspective other characteristics suggest that these mothers may be better situated than are the other groups (they are less likely to be in the youngest age category and they have more psychological resources at enrollment). Low attending mothers may have had less need for the program, and they and their nurses likely recognized this and, as they had been guided to do, adjusted the visitation schedule to reflect the fewer needs of the low attenders.

This pattern of results is consistent with an earlier analysis of total visits completed, which found that mothers with moderately high psychological resources had the fewest visits, whereas mothers with very high or very low psychological resources had more visits.¹² The outcomes for these mothers and their children were at least as good as those of high attenders, and they had fewer subsequent pregnancies than did their matched control group. This suggests that these families reaped the benefits from the program in the early months while they were attending, and a lower level of subsequent involvement at least was not harmful.

Increasing attenders, by contrast, had intake characteristics that were generally poorer than were those of the other groups. They had significantly less education and higher rates of preterm delivery and neonatal intensive care use. Unadjusted significance tests suggest that they may have lower psychological resources and may experience greater neighborhood poverty. Although this group received the second highest number of visits, their outcomes were worse than were those of the other groups. Children in this group were likely to have been perceived as being more vulnerable, which probably motivated both the nurse and the mother to complete visits after delivery. A longer hospital stay is likely to have delayed receipt of postpartum visits.

The low attendance pattern may reflect the nurses’ choosing to focus their limited time on those with premature babies, who were clearly in greater need. Although we were not able to determine how individual nurses managed their caseloads, we think that the risks of the increasing attender group were greater than those of the low attender group, and the nurses’ corresponding calculation of need likely accounts for the shift in patterns of visitation between these 2 groups. It is revealing, we think, that mothers in the increasing attender group reported better relationships with their nurses than did low attenders, whose needs and frequency of visits were lower, especially after delivery.

Implications

The identified classes suggest that a simple dose–response relationship does not best describe visit attendance in this program. Future studies examining visit attendance should consider patterns of attendance in addition to dose.

Highest grade completed may be a good predictor of attendance patterns because many factors contribute to school success, such as aptitude, stable environment, supportive family, and age. These factors may also be related to a mother’s visit attendance. Although we did not find any other characteristics to be significant, a larger sample size and different measures may provide additional valuable predictors.

Prediction of class membership during pregnancy and the early weeks following delivery may be a useful approach to managing nurse caseloads. A nurse who has many clients who miss appointments during pregnancy and eventually become increasing attenders may have more difficulty providing needed care to her families than does a nurse with low-risk clients, who may actually need fewer visits after delivery.

Caseload management is important not only to provide adequate care to families but also to prevent burnout and turnover among nurses. Participant attrition increases nurse turnover,³ along with hiring and training costs. Olds and his group are developing a tool for nurses to use that more comprehensively characterizes families’ strengths and risks (the Strength and Risk framework), which may help nurses understand families’ needs for the program and make adjustments in visit frequency that align with families’ perceptions of their needs.⁴⁵ Our findings provide empirical evidence to support use of this framework.

Limitations

The data we examined were from a single, closely supervised research trial site. Different NFP sites with fewer resources for supervision may find different visit attendance patterns. The specific patterns we found therefore may not reflect ongoing NFP practice. Moreover, the moderate sample size in this study led to classes with few participants, which limited statistical power. This is an issue that was particularly challenging for the analysis of achievement test outcomes, which employed only half of the sample.

It also is important to keep in mind that the data for this study came from a trial conducted in the early 1990s. Changes in technologies (such as cell phones) and sociocultural changes may also affect visit attendance in today’s environment. Replication of these analyses with currently operating NFP programs and with different programs would provide valuable information about visit dosage patterns and families’ needs and outcomes.

Conclusions

Studying visit pattern trajectories is likely to help us understand visit dosage, leading to better allocation of scarce visitor time. The exploration of interactions between nurses and mothers with respect to scheduling and attending visits is likely to improve home visiting programs in community practice.