This is a critical abstract of an economic evaluation that meets the criteria for inclusion on NHS EED. Each abstract contains a brief summary of the methods, the results and conclusions followed by a detailed critical assessment on the reliability of the study and the conclusions drawn.

Medicaid child health screenings using mailed pamphlets, phone calls and home visits.

Primary prevention - screening.

Cost-effectiveness analysis.

Children from birth to age 21 who are eligible for Medicaid, with no disabilities. The mean age of the children was 8.7 years.

The practice setting was primary care. The economic study was carried out in rural North Carolina, USA.

The effectiveness and resource data were taken from September 1990 - August 1992. Dollar values were based on 1993.

Single study.

Costing was undertaken alongside the effectiveness study using the same patient sample as that used in the effectiveness study.

The study included 2541 randomly selected families of which 488 did not meet the inclusion criteria. Sample selection of counties was based on 5 criteria:

(1) over half of the population resided in the county's rural areas;

(2) under 35% of all eligible children in the county obtained screenings in a 1-year period;

(3) driving time from the university research office was under 3 hours;

(4) there were no major local efforts to increase screening;

and (5) local officials wanted to collaborate in the study.

Power calculations showed that 330 families per study group would provide adequate statistical power to detect the difference between groups having screening rates of 25% vs 35%. With 10% over sampling to protect against data losses, 363 families per group were targeted. The total number of families in the control group with phone, was 298 and in the intervention group, 885. In the no-phone category, the control group totalled 300 and there was 570 in the intervention group.

Randomized controlled trial. It was multi-centred with six medically underserved counties. The duration of follow up of the cohort was for at least one targeted child in a family within 4 months after intervention or for control group, 4 months after the midpoint of the relevant sample's 3-months intervention period. There were no refusals of phone or home visit interventions among families reached. Pamphlets/letters reached 99% of with-phone families and 97% of no-phone families. Phone calls reached 57% of families with phones. Home visits reached 70% of with-phone families and 56% of no-phone families.

The analysis of the study was based on intention to treat. Main outcome was given by the number of success in producing health screening within the follow-up period. Comparability of groups was addressed.

After intervention, about 90% of each group(476 families) said they wanted to use the program and 212 (45%) asked for an appointment to be scheduled. Among families with phones, there were significant differences in effectiveness between informing methods. The control method produced the fewest screenings and the pamphlet/letter, phone call and home visit produced more screenings. Among families without phones, differences among the intervention and control methods were not significant. There were identical numbers of screenings produced by the pamphlet/letter and home visits for both methods. The p-value for with-phone families was P<.001 and the unadjusted odds ratio and 95% confidence intervals (shown in brackets) respectively, were as follows for the informing methods:

Pamphlet/letter1.40 (0.69, 2.85);

Phone call2.85 (1.50, 5.42);

Home visit3.95(2.13, 7.31).

The p-value for no-phone families was P=.229 and the unadjusted odds ratio and 95% confidence intervals (shown in brackets) respectively, are as follows for the informing methods:

Pamphlet/letter1.61 (0.85, 3.03);

Home visit1.67(0.88, 3.16).

The absolute effectiveness of the interventions in terms of producing screenings was low. The low effectiveness was due partly to many families who were targeted but not reached by an intervention. Such families obtained screenings at the same rate as families in the control groups. A subanalysis of families reached did not show higher rates for any intervention.

A stepwise logistic regression was conducted in order to identify the variables that might influence parents to obtain health screening for their children.

Production of additional health screening within the follow-up period.

Average costs were used. Quantities/costs were not analysed separately. The direct costs listed were borne by the health service. The dollar values were based on 1993 mid-point hourly salaries. Costs were based on actual data. They included materials, phone charges, travel, labour, etc.

US dollars ($).

No sensitivity analysis was carried out.

After-intervention, about 90% of each group(476 families) said they wanted to use the program and 212 (45%) asked for an appointment to be scheduled. Among families with phones, there were significant differences in effectiveness among informing methods. The control method produced the fewest screenings and the pamphlet/letter, phone call and home visit produced more screenings. Among families without phones, differences among the intervention and control methods were not significant. There were identical numbers of screenings produced by the pamphlet/letter and home visits for both methods.

The p-value for with-phone families was P<.001 and the unadjusted odds ratio and 95% confidence intervals (shown in brackets)respectively, are as follows for the informing methods:

Pamphlet/letter1.40 (0.69, 2.85);

Phone call2.85 (1.50, 5.42);

Home visit3.95(2.13, 7.31).

The p-value for no-phone families was P=.229 and the unadjusted odds ratio and 95% confidence intervals (shown in brackets)respectively, are as follows for the informing methods:

Pamphlet/letter1.61 (0.85, 3.03);

Home visit1.67(0.88, 3.16).

The average cost per intervention conducted by pamphlet/letter was $2.62, by phone call $9.33 and by home visit $31.68.

Relative to the control group, the incremental cost-effectiveness ratio (cost per additional family getting a Medicaid health screening) of with-phone families was $161 for pamphlet/letter, phone call $137 and home visit $306. For no-phone families by pamphlet/letter the ratio was $91 and home visit was $1022.

The study verified that briefly informing parents about the programme is unlikely to result in adequate use of the programme. Although the interventions produced more screening than the control method, the increases in absolute terms were minimal. Existing outreach methods should be improved and new ones developed and tested in order to be more cost-effective. Systems-level changes may be needed to enable families to use Medicaid health services. Therefore concurrent changes in the health care delivery system are needed.

(a) Overall the study can be termed good. Its study design, sample representation and trial methodology seemed to be sound.

(b) The authors could have provided more information on the cost analysis.

(c) The generalisability of these study results to other settings should be carefully evaluated.

Agency for Health Care Policy and Research Grant Number 5-R01-HS-06507