Research in Nursing & Health, 1990, 13,163 - 171
Medicaid Maternal and Child Health Care: Prepaid Plans vs. Private Fee-for-Service Janet Reis
A randomly selected sample of 98 inner-city new mothers was interviewed with regard to (a) their level of satisfaction with and perceived barriers to pediatric and maternal health care, and (b) the health status of their infants. The mothers were divided according to their Medicaid status: 39 (40%) were enrolled into a Medicaid prepaid plan (PPP) and 59 (60%) were enrolled as regular Medicaid fee-for-service (FFS) patients. Comparison of the two groups showed that infant health status was similar but that PPP mothers reported a higher incidence of untreated infant problems and perceived more barriers to pediatric care. Birth outcomes also were similar for the two groups but FFS mothers and infants were hospitalized for significantly longer periods postpartum. The implications of these findings for the delivery of maternal and child health services to low-income populations through the Medicaid program are discussed.
The Medicaid program remains the major source of health care for indigent women and children (Kasper, 1987). Pressures to contain health care costs, however, compel state governments to search for ways to reform Medicaid services (Brown & Cousineau, 1984). Of particular note are prepaid plans that require the assignment of a Medicaid recipient to one health care provider who manages the scope of the patient’s office- or clinic-based services, referrals to other providers, and hospital admissions. The provider in turn contracts with the state Medicaid program for a guaranteed but capitated payment (Freund, 1986). This report is a summary of Medicaid recipients’ perceptions of access and quality of services received through prepaid versus fee-for-service providers. The three areas of health care services chosen for study (family planning, obstetrics, and well child care) traditionally are regarded as key components of a comprehensive maternal and child health system (Wallace, Gold & Oglesby, 1982). Recent evidence shows that declines in public spending have limited the availability of these services to certain groups of low-income women
and children (Alan Guttmacher Institute, 1987; Butler, Winter, Singer, & Wenger, 1985; Rosenbaum & Johnson, 1986). The breadth of change in patient and provider roles represented by prepaid plans suggests that careful study is needed to determine if prepaid plans exacerbate or improve already existing problems with access to basic maternal and child health services for indigent patients. Levels of patient satisfaction, access to care, and other important dimensions have been researched extensively for white, middle-class families enrolled in prepaid plans (Berki & Ashcroft, 1980). Review of the literature on Medicaid HMOs shows, however, that the perspective of the Medicaid recipient is virtually ignored (Reis & Olson, 1987). Data on the HMO Medicaid experience are limited to trends in enrollmenVdisenrollment, and some data studies of dissatisfaction with the marketing techniques of prepaid plans (Health Care Financing Administration, 1988). Given the paucity of information on the consumer response to Medicaid prepaid plans, the results of this study offer some new insights as
Janet Reis, PhD, is a senior research analyst in the College of Nursing, University of Illinois at Urbana-Champaign. This research was supported in part by a grant from the Ford Foundation to the Center for Health Services and Policy Research, Northwestern University, Evanston, Illinois. This article was received on June 19, 1989, was revised, and accepted for publication December 1, 1989. Requests for reprints can be addressed to Dr. Janet Reis, College of Nursing, University of Illinois at Urbana-Champaign, 11 15% W. Oregon Street, Urbana, IL 61801.
0 1990 John Wilry & Sons. Inc. 0160-6891/90/030163-09$04.00
164
RESEARCH IN NURSING & HEALTH
to the workings of these systems in certain communities. The community in which the study was conducted has been selected by the Illinois Department of Public Aid (IDPA) as a demonstration site for Medicaid prepaid plans. The neighborhood has been identified as one of the poorest in the nation, with a median income of $7,600 in 1980 ($3,700 in constant 1967 dollars; de Vise, 1984). High rates of hospital emergency room utilization and prescription drugs within the community’s Medicaid population led IDPA to experiment with an alternative Medicaid delivery system.
METHOD Sample The sample of 98 new mothers was selected from a group of Medicaid participants not registered with the Department of Health (DOH) and known to be eligible for participation in a Medicaid HMO because they resided in IDPAs Medicaid demonstration area. The final sample of 98 mothers ranged in age from 15 to 41 years ( M = 23.2, SD = 5.7) and divided between 39 PPP and 59 FFS mother-infant pairs. The two groups of mothers did not differ from each other in employment status (67% unemployed) or in length of time living at current residence (M = 10 years, SD = 7.2). PPP mothers were more likely than FFS mothers to have been married or to be currently married (33% PPP mothers versus 10% FFS mothers; x2 (4) = 10.35, p =z .03). All mothers reported an annual income of less than $8,000 per year.
instrument The 27-page interview instrument was based on the survey used in a national survey of utilization of health care in the United States (Aday, Anderson, & Fleming, 1980). Two major domains were addressed in the adapted survey instrument: (a) use of pediatric services for specific conditions and level of satisfaction with care received; (b) the mother’s labor and delivery experience and postnatal practice of family planning. In addition, the mother’s experience with the health department, the reasons she had not registered the infant there, her perception of the baby’s health, and the degree of worry it caused her were assessed. The original survey instrument was reviewed by a five-person committee composed of 2 uni-
versity-based researchers and three DOH employees who had a combined total of 48 years working in the community. The committee first selected questionnaire items falling into the areas of interest. Each question then was reviewed for appropriateness of phrasing and terminology. Terms referring to the health care system (e.g., Medicaid) and specific health conditions (e.g., conjunctivitis) were changes to correspond to the words commonly used by community residents (e.g., Green Card and cold in eyes). The revised instrument then was pilot tested with 10 community residents to assess understandability of the questionnaire items. Information on the psychometic properties for the original survey instrument is not reported (Aday et al., 1980). In view of the extensive national use of the survey, however, it is assumed that the questions had been scrutinized for content and construct validity. Revisions made for the current study also were reviewed for these two facets of validity. Reliability was assessed using Cronbach’s alpha for the 9-item satisfaction with care subscale (alpha = .75; see Table 1 for a listing of items).
Procedure The initial universe of 1,488 1985 births was identified through public health nursing records. Forty-four percent (656) of the infants were registered with DOH as determined by review of clinic patient registration logs. For purposes of this study, patients who attended a local department of health clinic were not eligible to participate because of the constraints in Medicaid eligibility imposed by the state public aid program on DOH patients, thus making their experience with the Medicaid program atypical. The remaining 832 infants were not registered as of November 1986 at the clinics serving the community area. The possibility remains that these infants were registered at one of the other 18 DOH facilities, but this was not verified. A sample of 200 infants was originally established for contact for a home interview by selecting every fourth record from the sample of 832 infants. Because of the subject recruitment difficulties discussed below, two additional samples were pulled, for a final sampling of 333 (47%) of the original population. All potential participants were sent a letter from a local community organization describing the purpose of the study. Efforts were then made to reach the mother by telephone. If the mother agreed to be interviewed, an appointment was scheduled at her convenience in her
MATERNAL AND C H I L D HEALTH CARE I R E I S
home. Prior to the beginning of the interview, the mother was read a letter of consent approved by the university’s institutional review board and the mother and interviewer co-signed the statement, indicating that they both understood the terms of participation. The three most frequently occurring reasons for nonparticipation of 285 mother-infant pairs initially identified through the DOH records were as follows: family moved out of area (72 cases), family could not be located by telephone or home visit (71 cases), and current registration with a DOH clinic (49 cases). Other reasons for nonparticipation included wrong phone number (22 cases), failure to find mother through family member (20 cases), infant death (8 cases), mother refused interview (8 cases) and broken interview appointment (4 cases). Three interviewers conducted the in-home interviews. Each interviewer was oriented to the questionnaire prior to pilot testing. A minimum of three tape-recorded pilot interviews were completed and reviewed to ensure completeness and appropriateness of questioning. Thirty-three of the interview tapes also were compared with the completed questionnaire to double check on the accuracy and completeness of the recorded information. Interviewers were required to make three attempts to contact the mother, including telephone call, visits to the purported residence, and contacts with relatives and neighbors who might be aware of the mother’s whereabouts. The majority of the interviews (67%) were conducted by one interviewer who had worked in the community for the past 15 years. As the study progressed, it became clear that of the three interviewers originally trained, this individual had the best rapport with the mothers and was most comfortable travelling through the neighborhood, which was known for its high crime rate.
RESULTS The results of the survey are presented in three sections corresponding to the maternal and child health services chosen for study. A synopsis of satisfaction with health care service concludes the results section.
Utilization of Family Planning The mothers were asked to summarize their family planning practice before and after the birth of the
165
index child. Fifty-seven percent (56) of the mothers used no method of family planning prior to becoming pregnant. Sixty-six percent (65) of the mothers had obtained some method of birth control since their child’s birth, but only 45% reported using any method of contraception at the time of the interview. T h q - o n e separate sources of family planning methods were named; for those women using some method, birth control pills and sponges were the most frequently used methods (1 3% and 21% respectively). No differences in patterns of utilization of family planning services emerged according to the mother’s Medicaid status.
Utilization of Perinatal Services A total of 24 hospitals were used for labor and delivery. Prior use by other family members was a more important reason for selection of site of delivery for the FFS group (x2(1) = 3.17, p .07 with Yates correction). The mothers did not differ in their rating of the importance of physician referral (39% overall), belief that the hospital offered good obstetrical care (21% overall), or knowledge of what obstetrical services were available at the facility (28% overall). FFS mothers were proportionately more satisfied with their hospitalization experience than the PPP group (76% versus 67%). The mothers did not differ in the rate of C-section (15% overall), the number of early deliveries (46% overall), nor the number of newborn complications (respiratory, jaundice and congenital abnormalities). Infant birth weight averaged 6.4 pounds. Despite similarities in birth outcomes, infants born to mothers in a FFS plan were somewhat more likely to be placed in an incubator than were infants born to PPP enrollees (20% versus 10%). FFS infants spent a significantly greater number of days in an incubator ( M = 9.7, SD = 5.90) than PPP infants ( M = 7.75, SD = 3.13), ((96) = 1.24, p G .05, one-tailed. FFS mothers also had longer hospital stays ( M = 4.37, SD = 3.6) than the comparison group ( M = 3.44, SD = 2.89), t(96) = 1.36, p .09, one-tailed.
Utilization of Pediatric Care Patterns of utilization of pediatric care were first assessed with mode of transportation and appointment making behavior. Walking was the most common mode of transportation for both groups of mothers (40% overall). FFS mothers were more likely to travel by car (27% versus 10%)and less
166
RESEARCH IN NURSING & HEALTH
likely to rely on public transportation (22% versus 39%) than were PPP mothers. The groups were identical in their use of walk-in visits (39%) and waiting times in physicidclinic office (17.5 min), but PPP mothers spent less time with their provider (16.6 min) than FFS mothers (21.6 min), r(95) = 1.99, p C .02, one-tailed. Utilization of pediatric care also was assessed with the mother's reports of the occurrence of common child health problems, use of health services for these problems and immunization history. The mothers were first presented with a list of 17 health problems that commonly occur for very young children and asked first if the child had ever had the problem and, if so, how many times the problem had occurred in the child's first year of life. The mothers were then asked if they had ever taken their child to the doctor for the problem and, if so, the number of times they had done so. Prepaid plan mothers were more likely than FFS mothers to report that their child had not had problems with slow weight gain, listlessness, and cold in eyes, although these differences were not statistically significant. The mothers gave comparable reports of their infants being free from weight loss (77% overall), diarrhea (33% overall), constipation (67% overall), vomiting (66% overall),
8o
fever (64% overall), ear infections (65% overall), white patches in mouth (86% overall), diaper rash (75% overall), coughing (64% overall), foot problems (94% overall), leg problems (95% overall), and developmental problems (92% overall). A significant difference appeared in the self-report of the incidence of wheezing, with 90% of the PPP mothers reporting no problem and 76% of the FFS mothers reporting no problem, ~ ' ( 1 )= 6 . 9 9 , ~4 .025. A comparison of the two groups of infants on these same health problems was made according to the occurrence of problems receiving no physician care as a percent of that problem receiving at least one physician visit in the first year of life. As shown in Figure 1, a larger percent of babies enrolled in a prepaid Medicaid program were reported by their mothers as receiving no medical attention for diarrhea, constipation, vomiting, cold in eyes, and diaper rash. Based on calculation from actual frequency of occurrence, a statistically significant association between Medicaid status and treatment status was observed for diarrhea, ~ ' ( 1 ) = 11.75, p C .001; vomiting, x2(1) = 11.00, p s .001; and cold in (runny) eyes, ~ ' ( 1 ) = 6.36, p S .025. No differences were observed between the two groups of infants in
3 rn
.-C a c3)
E
0
3
.-Q *-'
u)
C 0 0
0,
C .-c .E
0
>
PPP FFS
> .c 0
C
.-3 C 0 0
FIGURE 1. Infant health problems receiving no physician care as percent of problems receiving at least one physician visit according to Medicaid status.
MATERNAL AND CHILD HEALTH CARE / RElS
167
100 O/O
80
within 2 months AClP on schedule wit1I AClP
60 40
20 0
FIGURE 2.
Percent of infants receiving immunizations according to AClP standards and Medicaid status.
terms of rates of hospitalization in the first year of life (14% hospitalized overall) or need for some type of surgery 3% overall). The mothers also were queried as to their children’s immunization status for diphtheria-pertussistetanus (DPT), oral polio vaccine (OPV) and measles-mumps (MM). Immunization status was confirmed with review of the actual record for 80% of the mothers; there was no difference between the two groups in terms of presentation of the record and completion of any of the immunization series. The percent of infants receiving these vaccines according to the schedule published by the Immunization Practices Advisory Committee (ACIP) of the U . S . Public Health Service (U.S. Department of Health and Human Services, 1986) is presented in Figure 2. The percentages displayed in the schedule show that more infants enrolled in a prepaid plan received their DPT2 and DPT3, OPV2, and MM vaccines at the recommended time than did infants taken to a fee-for-service provider. Prepaid enrollees overall had a greater percent of infants immunized within 2 months of the recommended schedule for DPTl(90% versus 72%), OPVl (87% versus 76%). and OPV3 (71% versus 63%). When asked how they would rate their child’s health in the first year of life, 28% of PPP mothers and 61% of FFS mothers selected “excellent” as a description; 69% and 31% of these two groups
respectively used the tern “good,” ~ ’ ( 3 )= 14.63, p < .005. As a group, PPP mothers were more likely than FFS mothers to state that they worried somewhat about their infant’s health, although the difference was not statistically significant. Seventy-four percent of the PPP group reported worrying somewhat to a great deal, in comparison with 66% of the FFS group.
Satisfaction with Care The majority of mothers (59% PPP and 64% FFS) cited good quality of care received as the most important feature of their health care. Dissatisfaction with health care services received was explained primarily by four variables for those mothers who responded to this question. In descending order of importance, these factors were long waiting time (39% PPP versus 36% FFS), no after service hours (13% PPP versus 17% FFS), long distance from home (21 % PPP versus 12% FFS). and poor quality care (8% PPP versus 2% FFS). More specific facets of satisfaction with the pediatric care received by their infants were assessed with nine questions addressing dimensions of care (see Table 1). When the response categories of “very dissatisfied and somewhat dissatisfied” were combined, more PPP mothers than FFS
Table 1. Satisfaction with Pediatric Care according to Medicaid Status Medicaid HMO
Medicaid fee-for-service (n = 59)
(n = 39) n Amount of time to go to health care provider Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Amount of time waited between scheduling and making visit Very dissatisfied Somewhat dissatisfied Neutral Satified Very satisfied Amount of waiting time in office or clinic Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Information from physician Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Cost of visit Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Quality of care Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Nurseiphysician concern Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Doctor’s willingness to make referral Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Care provided overall Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied
(Yo)
n
2. 3 10 5 19
2 1 11 7 38
3 3 5 13 14
1 4 13 15 26
4 3 13 6 13
6 8 16 12 17
3 4 3 8 21
1 1 3 12 42
1 0 2 1 35
0 0 3 3 53
2 1 3 10 23
0 1 3 11 4
3 0 3 8 25
1 2 6 6 44
3 2 2 4 27
2 1 8 7 39
1 1 5 7 25
0 1 7
5 46
(Oh
1
169
MATERNAL AND CHILD HEALTH CARE / RElS
50
1
40 -
30 -
H
ea
20 10 0
PPP FFS
7-
T
Hlth Dept
MD
L HMO
-r
Hosp
ER
Preference for Provider FIGURE 3. Mother’s preferred health care provider according to Medicaid status.
of sample), and hospital’s refusal to accept Medmothers were displeased with the amount of time icaid patients (83% of sample). The majority of required to go to health provider (13% versus the sample (80%)also felt it was difficult to obtain 5 % ) , information received from physician (7% Medicaid and therefore difficult to obtain health versus 2%), quality of care (8% versus 2%), and care. amount of concern expressed by nurseiphysician The mother’s preferred choice of health care (8% versus 5%). More FFS mothers than PPP provider according to her current Medicaid status mothers were dissatisfied with waiting time (24% is summarized in Figure 3. The mothers differed versus 18%). None of the differences observed in their choice of provider, with more PPP mothers between the two groups of mothers were found (41%) than FFS mothers (22%) desiring to go to to be statistically significant. Following these questions on level of satisfacthe department of health, x2(5) = 11.08, p s .05. In comparison with the other group, FFS tion, the respondents were asked if they perceived any barriers to their being able to obtain the care mothers were more interested in private physicians (36% versus 31%), hospital clinics (25% versus they felt their child needed. Significantly more 13%) and hospital emergency rooms (17% versus PPP mothers (21%) than FFS mothers ( 5 % ) in10%). In terms of past experience, 87% of the dicated that there had been a time when they PPP mothers and 70% of the FFS mothers had thought their baby needed a doctor but they could ever used CDOH services. Twenty-eight percent not get access to health care, ~ ‘ ( 2 )= 6.14, p S of the PPP group compared with 8% of the FFS .05. PPP mothers indicated that they had enmothers, stated that DOH services must be paid countered more specific barriers to care than had for, x2(3) = 7.79, p s .05. FFS mothers (21 positive responses versus 7). Being restricted to one physician by Public Aid was marginally significant, ~ ’ ( 1 ) = 3.31, p 6 .07 with Yates correction. DISCUSSION PPP mothers also were more likely to state that people on Public Aid do not receive as much good care as people not enrolled in the program, ~ ~ (=2 ) The results of the survey of inner-city mothers identified a number of differences in health status, 4.83, p S .09. In aggregate, the mothers agreed maternal perception of access to health care, and that difficulties in access to care were created with preference for health care provider according to enrollment Medicaid HMOs (68% of sample), the type of Medicaid program in which the mothers restrictions to one proyider (74% of sample), prowere enrolled. A discussion of the limitations of vider’s refusal to accept Medicaid patients (84%
170
RESEARCH IN NURSING & HEALTH
these findings is followed by suggestions of issues for further study. The generalizability of the survey results is limited by the relatively small sample size. Although the mothers interviewed were representative of the population of community women enrolled in Medicaid in terms of age, race and marital status, a larger sample would be desirable. The project team had originally planned to conduct interviews with more mothers, but were unable to meet the target sample size because of resource limitations and unanticipated difficulties in finding families. Another factor potentially limiting the validity of the survey findings is reliance on the mother as the primary source of information on health status. Ideally, the mother’s statements would have been checked against her medical record. Prepaid Medicaid providers do not, however, currently permit outside organizations to inspect their clinical documentation. In lieu of cross-checking the mother’s recall or illness and treatment with the full medical report, a check was done with the infant’s immunization record. This cross-validation documented a high level of accuracy in maternal reporting paralleling the findings of other studies of parental reports of their children’s illnesses (Bergner & Rothman, 1987). Inaccuracies in the mothers’ impressions of the accessibility and quality of their health care represents another source of bias more difficult to address than self-reports of health status. Affirmation of the respondents’ perceptions of their providers would require evidence on administrative policies and management procedures, as well as objective review of the quality of care provided. Assembling such information was beyond the scope of this project. A number of differences between the prepaid and fee-for-service experiences related to clinical and administrative practices delineate issues for future study as discussed below. From a patient care perspective, the survey findings suggest that a comparative evaluation be done of specific clinical protocols employed by providers in the various Medicaid models. Newborns enrolled in a prepaid plan had apparently contradictory experiences in that their mothers reported a lower use of newborn hospital services and primary care for infant illnesses but more \ complete and timely use of immunization services. More information is needed to determine if different models promote systematic variations in the quality of patient care. In a similar vein, objective assessments of the availability of services in the prepaid and fee-for-service Medicaid models are in order. The PPP mothers’ descriptions of difficulties in obtaining certain services may be related
to clinic management policies or may reflect some other set of factors outside the realm of the health care system. Such recommendations for future studies must be reviewed in the context of the debate on Medicaid reform and the underinsured (Wilensky, 1987). The State of Illinois Department of Public Aid has recently issued guidelines for the Medicaid program that include expanding services for pregnant women and children up to 185% of the federal poverty level (Illinois Department of Public Aid, 1989). At the same time, IDPA must find ways to contain Medicaid expenses. Exploration of clinical and administrative practices in different Medicaid models would provide information needed to strike a judicious balance between accessibility, quality, and costs for the population needing maternal and child health services.
REFERENCES Aday, L.A., Anderson, R., & Fleming, G.V. (1980). Health care in the United States: Equitable f o r whom? Beverly Hills: Sage. Alan Guttmacher Institute (1987). Blessed evenfs and the bottom line: Financing maternity care in the United States. New York: Alan Guttmacher Institute. Bergner, M . , & Rothman, M.L. (1987). Health status measures: An overview and guide for selection. Annual Review of Public Health. 8 , 191-210. Berki, S . E . , & Ashcroft. M.L.F. (1980). HMO enrollment: Who joins what and why: A review of the literature. Milbank Memorial Fund Quarterlv, 58, 588-632. Brown, E.R., & Cousineau, M.R. (1984). Effectiveness of state mandates to maintain local government health services for the poor. Journal of Health Politics, Policv and Law, 9, 223-235. Butler, J.A., Winter, W.D., Singer, J.P., & Wenger, M. (1985). Medical care use and expenditure among children and youth in the United States: Analysis of a national probability sample. Pediatrics. 7 6 , 495-507. de Vise, P. (1984). The nation’s sixteen poorest neighborhoods (Chicago regional inventory, Working Paper No. 1177). City Club of Chicago. Freund, D.A. (1986). The private delivery of Medicaid services: Lessons for administrators, providers, and policymakers. Journal ofAmbulatory Care Management, 9, 54-65. Health Care Financing Administration (1988). Medicaid and prepayment: A bibliographic essay. Washington, DC: U . S . Department of Health and Human Services, Public Health Service. Illinois Department of Public Aid (1989). A proposal for Medicaid reform: Carrying the challenge into
MATERNAL AND CHILD HEALTH CARE 1 RElS
the future. Springfield, 1L: Illinois Department of Public Aid. Kasper, J . D . (1987). The importance of type of usual source of care for children’s physician access and expenditures. Medical Care, 2 5 , 386-398. Reis, J . . & Olson, L. (1987). The Medicaid program and consumer needs: A survey among residents of a poor Chicago neighborhood. Public Health Reports, 102, 530-538. Rosenbaum, S . , &Johnson. K . (1986). Providing health care for low income children: Reconciling child health goals with child health financing
171
realities. The Milbank Quarterly, 6 4 , 442-478. U . S . Department of Health and Human Services, Centers for Disease Control (1986). New recommended schedule for active immunization of normal infants and children. Morbidity andMortality Weekly Report, 35(37), 577-579. Wallace, H . M . , Gold, E.M., & Oglesby, A.C. (1982). Maternal and child health practice: Problems, resources and methods of delivery (2nd ed.). New York: Wiley. Wilensky, G.R. (1987). Viable strategies for dealing with the underinsured. Health Affairs, 6, 33-46.
Medicaid Maternal and Child Health Care: Prepaid Plans vs. Private Fee-for-Service Janet Reis
A randomly selected sample of 98 inner-city new mothers was interviewed with regard to (a) their level of satisfaction with and perceived barriers to pediatric and maternal health care, and (b) the health status of their infants. The mothers were divided according to their Medicaid status: 39 (40%) were enrolled into a Medicaid prepaid plan (PPP) and 59 (60%) were enrolled as regular Medicaid fee-for-service (FFS) patients. Comparison of the two groups showed that infant health status was similar but that PPP mothers reported a higher incidence of untreated infant problems and perceived more barriers to pediatric care. Birth outcomes also were similar for the two groups but FFS mothers and infants were hospitalized for significantly longer periods postpartum. The implications of these findings for the delivery of maternal and child health services to low-income populations through the Medicaid program are discussed.
The Medicaid program remains the major source of health care for indigent women and children (Kasper, 1987). Pressures to contain health care costs, however, compel state governments to search for ways to reform Medicaid services (Brown & Cousineau, 1984). Of particular note are prepaid plans that require the assignment of a Medicaid recipient to one health care provider who manages the scope of the patient’s office- or clinic-based services, referrals to other providers, and hospital admissions. The provider in turn contracts with the state Medicaid program for a guaranteed but capitated payment (Freund, 1986). This report is a summary of Medicaid recipients’ perceptions of access and quality of services received through prepaid versus fee-for-service providers. The three areas of health care services chosen for study (family planning, obstetrics, and well child care) traditionally are regarded as key components of a comprehensive maternal and child health system (Wallace, Gold & Oglesby, 1982). Recent evidence shows that declines in public spending have limited the availability of these services to certain groups of low-income women
and children (Alan Guttmacher Institute, 1987; Butler, Winter, Singer, & Wenger, 1985; Rosenbaum & Johnson, 1986). The breadth of change in patient and provider roles represented by prepaid plans suggests that careful study is needed to determine if prepaid plans exacerbate or improve already existing problems with access to basic maternal and child health services for indigent patients. Levels of patient satisfaction, access to care, and other important dimensions have been researched extensively for white, middle-class families enrolled in prepaid plans (Berki & Ashcroft, 1980). Review of the literature on Medicaid HMOs shows, however, that the perspective of the Medicaid recipient is virtually ignored (Reis & Olson, 1987). Data on the HMO Medicaid experience are limited to trends in enrollmenVdisenrollment, and some data studies of dissatisfaction with the marketing techniques of prepaid plans (Health Care Financing Administration, 1988). Given the paucity of information on the consumer response to Medicaid prepaid plans, the results of this study offer some new insights as
Janet Reis, PhD, is a senior research analyst in the College of Nursing, University of Illinois at Urbana-Champaign. This research was supported in part by a grant from the Ford Foundation to the Center for Health Services and Policy Research, Northwestern University, Evanston, Illinois. This article was received on June 19, 1989, was revised, and accepted for publication December 1, 1989. Requests for reprints can be addressed to Dr. Janet Reis, College of Nursing, University of Illinois at Urbana-Champaign, 11 15% W. Oregon Street, Urbana, IL 61801.
0 1990 John Wilry & Sons. Inc. 0160-6891/90/030163-09$04.00
164
RESEARCH IN NURSING & HEALTH
to the workings of these systems in certain communities. The community in which the study was conducted has been selected by the Illinois Department of Public Aid (IDPA) as a demonstration site for Medicaid prepaid plans. The neighborhood has been identified as one of the poorest in the nation, with a median income of $7,600 in 1980 ($3,700 in constant 1967 dollars; de Vise, 1984). High rates of hospital emergency room utilization and prescription drugs within the community’s Medicaid population led IDPA to experiment with an alternative Medicaid delivery system.
METHOD Sample The sample of 98 new mothers was selected from a group of Medicaid participants not registered with the Department of Health (DOH) and known to be eligible for participation in a Medicaid HMO because they resided in IDPAs Medicaid demonstration area. The final sample of 98 mothers ranged in age from 15 to 41 years ( M = 23.2, SD = 5.7) and divided between 39 PPP and 59 FFS mother-infant pairs. The two groups of mothers did not differ from each other in employment status (67% unemployed) or in length of time living at current residence (M = 10 years, SD = 7.2). PPP mothers were more likely than FFS mothers to have been married or to be currently married (33% PPP mothers versus 10% FFS mothers; x2 (4) = 10.35, p =z .03). All mothers reported an annual income of less than $8,000 per year.
instrument The 27-page interview instrument was based on the survey used in a national survey of utilization of health care in the United States (Aday, Anderson, & Fleming, 1980). Two major domains were addressed in the adapted survey instrument: (a) use of pediatric services for specific conditions and level of satisfaction with care received; (b) the mother’s labor and delivery experience and postnatal practice of family planning. In addition, the mother’s experience with the health department, the reasons she had not registered the infant there, her perception of the baby’s health, and the degree of worry it caused her were assessed. The original survey instrument was reviewed by a five-person committee composed of 2 uni-
versity-based researchers and three DOH employees who had a combined total of 48 years working in the community. The committee first selected questionnaire items falling into the areas of interest. Each question then was reviewed for appropriateness of phrasing and terminology. Terms referring to the health care system (e.g., Medicaid) and specific health conditions (e.g., conjunctivitis) were changes to correspond to the words commonly used by community residents (e.g., Green Card and cold in eyes). The revised instrument then was pilot tested with 10 community residents to assess understandability of the questionnaire items. Information on the psychometic properties for the original survey instrument is not reported (Aday et al., 1980). In view of the extensive national use of the survey, however, it is assumed that the questions had been scrutinized for content and construct validity. Revisions made for the current study also were reviewed for these two facets of validity. Reliability was assessed using Cronbach’s alpha for the 9-item satisfaction with care subscale (alpha = .75; see Table 1 for a listing of items).
Procedure The initial universe of 1,488 1985 births was identified through public health nursing records. Forty-four percent (656) of the infants were registered with DOH as determined by review of clinic patient registration logs. For purposes of this study, patients who attended a local department of health clinic were not eligible to participate because of the constraints in Medicaid eligibility imposed by the state public aid program on DOH patients, thus making their experience with the Medicaid program atypical. The remaining 832 infants were not registered as of November 1986 at the clinics serving the community area. The possibility remains that these infants were registered at one of the other 18 DOH facilities, but this was not verified. A sample of 200 infants was originally established for contact for a home interview by selecting every fourth record from the sample of 832 infants. Because of the subject recruitment difficulties discussed below, two additional samples were pulled, for a final sampling of 333 (47%) of the original population. All potential participants were sent a letter from a local community organization describing the purpose of the study. Efforts were then made to reach the mother by telephone. If the mother agreed to be interviewed, an appointment was scheduled at her convenience in her
MATERNAL AND C H I L D HEALTH CARE I R E I S
home. Prior to the beginning of the interview, the mother was read a letter of consent approved by the university’s institutional review board and the mother and interviewer co-signed the statement, indicating that they both understood the terms of participation. The three most frequently occurring reasons for nonparticipation of 285 mother-infant pairs initially identified through the DOH records were as follows: family moved out of area (72 cases), family could not be located by telephone or home visit (71 cases), and current registration with a DOH clinic (49 cases). Other reasons for nonparticipation included wrong phone number (22 cases), failure to find mother through family member (20 cases), infant death (8 cases), mother refused interview (8 cases) and broken interview appointment (4 cases). Three interviewers conducted the in-home interviews. Each interviewer was oriented to the questionnaire prior to pilot testing. A minimum of three tape-recorded pilot interviews were completed and reviewed to ensure completeness and appropriateness of questioning. Thirty-three of the interview tapes also were compared with the completed questionnaire to double check on the accuracy and completeness of the recorded information. Interviewers were required to make three attempts to contact the mother, including telephone call, visits to the purported residence, and contacts with relatives and neighbors who might be aware of the mother’s whereabouts. The majority of the interviews (67%) were conducted by one interviewer who had worked in the community for the past 15 years. As the study progressed, it became clear that of the three interviewers originally trained, this individual had the best rapport with the mothers and was most comfortable travelling through the neighborhood, which was known for its high crime rate.
RESULTS The results of the survey are presented in three sections corresponding to the maternal and child health services chosen for study. A synopsis of satisfaction with health care service concludes the results section.
Utilization of Family Planning The mothers were asked to summarize their family planning practice before and after the birth of the
165
index child. Fifty-seven percent (56) of the mothers used no method of family planning prior to becoming pregnant. Sixty-six percent (65) of the mothers had obtained some method of birth control since their child’s birth, but only 45% reported using any method of contraception at the time of the interview. T h q - o n e separate sources of family planning methods were named; for those women using some method, birth control pills and sponges were the most frequently used methods (1 3% and 21% respectively). No differences in patterns of utilization of family planning services emerged according to the mother’s Medicaid status.
Utilization of Perinatal Services A total of 24 hospitals were used for labor and delivery. Prior use by other family members was a more important reason for selection of site of delivery for the FFS group (x2(1) = 3.17, p .07 with Yates correction). The mothers did not differ in their rating of the importance of physician referral (39% overall), belief that the hospital offered good obstetrical care (21% overall), or knowledge of what obstetrical services were available at the facility (28% overall). FFS mothers were proportionately more satisfied with their hospitalization experience than the PPP group (76% versus 67%). The mothers did not differ in the rate of C-section (15% overall), the number of early deliveries (46% overall), nor the number of newborn complications (respiratory, jaundice and congenital abnormalities). Infant birth weight averaged 6.4 pounds. Despite similarities in birth outcomes, infants born to mothers in a FFS plan were somewhat more likely to be placed in an incubator than were infants born to PPP enrollees (20% versus 10%). FFS infants spent a significantly greater number of days in an incubator ( M = 9.7, SD = 5.90) than PPP infants ( M = 7.75, SD = 3.13), ((96) = 1.24, p G .05, one-tailed. FFS mothers also had longer hospital stays ( M = 4.37, SD = 3.6) than the comparison group ( M = 3.44, SD = 2.89), t(96) = 1.36, p .09, one-tailed.
Utilization of Pediatric Care Patterns of utilization of pediatric care were first assessed with mode of transportation and appointment making behavior. Walking was the most common mode of transportation for both groups of mothers (40% overall). FFS mothers were more likely to travel by car (27% versus 10%)and less
166
RESEARCH IN NURSING & HEALTH
likely to rely on public transportation (22% versus 39%) than were PPP mothers. The groups were identical in their use of walk-in visits (39%) and waiting times in physicidclinic office (17.5 min), but PPP mothers spent less time with their provider (16.6 min) than FFS mothers (21.6 min), r(95) = 1.99, p C .02, one-tailed. Utilization of pediatric care also was assessed with the mother's reports of the occurrence of common child health problems, use of health services for these problems and immunization history. The mothers were first presented with a list of 17 health problems that commonly occur for very young children and asked first if the child had ever had the problem and, if so, how many times the problem had occurred in the child's first year of life. The mothers were then asked if they had ever taken their child to the doctor for the problem and, if so, the number of times they had done so. Prepaid plan mothers were more likely than FFS mothers to report that their child had not had problems with slow weight gain, listlessness, and cold in eyes, although these differences were not statistically significant. The mothers gave comparable reports of their infants being free from weight loss (77% overall), diarrhea (33% overall), constipation (67% overall), vomiting (66% overall),
8o
fever (64% overall), ear infections (65% overall), white patches in mouth (86% overall), diaper rash (75% overall), coughing (64% overall), foot problems (94% overall), leg problems (95% overall), and developmental problems (92% overall). A significant difference appeared in the self-report of the incidence of wheezing, with 90% of the PPP mothers reporting no problem and 76% of the FFS mothers reporting no problem, ~ ' ( 1 )= 6 . 9 9 , ~4 .025. A comparison of the two groups of infants on these same health problems was made according to the occurrence of problems receiving no physician care as a percent of that problem receiving at least one physician visit in the first year of life. As shown in Figure 1, a larger percent of babies enrolled in a prepaid Medicaid program were reported by their mothers as receiving no medical attention for diarrhea, constipation, vomiting, cold in eyes, and diaper rash. Based on calculation from actual frequency of occurrence, a statistically significant association between Medicaid status and treatment status was observed for diarrhea, ~ ' ( 1 ) = 11.75, p C .001; vomiting, x2(1) = 11.00, p s .001; and cold in (runny) eyes, ~ ' ( 1 ) = 6.36, p S .025. No differences were observed between the two groups of infants in
3 rn
.-C a c3)
E
0
3
.-Q *-'
u)
C 0 0
0,
C .-c .E
0
>
PPP FFS
> .c 0
C
.-3 C 0 0
FIGURE 1. Infant health problems receiving no physician care as percent of problems receiving at least one physician visit according to Medicaid status.
MATERNAL AND CHILD HEALTH CARE / RElS
167
100 O/O
80
within 2 months AClP on schedule wit1I AClP
60 40
20 0
FIGURE 2.
Percent of infants receiving immunizations according to AClP standards and Medicaid status.
terms of rates of hospitalization in the first year of life (14% hospitalized overall) or need for some type of surgery 3% overall). The mothers also were queried as to their children’s immunization status for diphtheria-pertussistetanus (DPT), oral polio vaccine (OPV) and measles-mumps (MM). Immunization status was confirmed with review of the actual record for 80% of the mothers; there was no difference between the two groups in terms of presentation of the record and completion of any of the immunization series. The percent of infants receiving these vaccines according to the schedule published by the Immunization Practices Advisory Committee (ACIP) of the U . S . Public Health Service (U.S. Department of Health and Human Services, 1986) is presented in Figure 2. The percentages displayed in the schedule show that more infants enrolled in a prepaid plan received their DPT2 and DPT3, OPV2, and MM vaccines at the recommended time than did infants taken to a fee-for-service provider. Prepaid enrollees overall had a greater percent of infants immunized within 2 months of the recommended schedule for DPTl(90% versus 72%), OPVl (87% versus 76%). and OPV3 (71% versus 63%). When asked how they would rate their child’s health in the first year of life, 28% of PPP mothers and 61% of FFS mothers selected “excellent” as a description; 69% and 31% of these two groups
respectively used the tern “good,” ~ ’ ( 3 )= 14.63, p < .005. As a group, PPP mothers were more likely than FFS mothers to state that they worried somewhat about their infant’s health, although the difference was not statistically significant. Seventy-four percent of the PPP group reported worrying somewhat to a great deal, in comparison with 66% of the FFS group.
Satisfaction with Care The majority of mothers (59% PPP and 64% FFS) cited good quality of care received as the most important feature of their health care. Dissatisfaction with health care services received was explained primarily by four variables for those mothers who responded to this question. In descending order of importance, these factors were long waiting time (39% PPP versus 36% FFS), no after service hours (13% PPP versus 17% FFS), long distance from home (21 % PPP versus 12% FFS). and poor quality care (8% PPP versus 2% FFS). More specific facets of satisfaction with the pediatric care received by their infants were assessed with nine questions addressing dimensions of care (see Table 1). When the response categories of “very dissatisfied and somewhat dissatisfied” were combined, more PPP mothers than FFS
Table 1. Satisfaction with Pediatric Care according to Medicaid Status Medicaid HMO
Medicaid fee-for-service (n = 59)
(n = 39) n Amount of time to go to health care provider Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Amount of time waited between scheduling and making visit Very dissatisfied Somewhat dissatisfied Neutral Satified Very satisfied Amount of waiting time in office or clinic Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Information from physician Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Cost of visit Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Quality of care Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Nurseiphysician concern Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Doctor’s willingness to make referral Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied Care provided overall Very dissatisfied Somewhat dissatisfied Neutral Satisfied Very satisfied
(Yo)
n
2. 3 10 5 19
2 1 11 7 38
3 3 5 13 14
1 4 13 15 26
4 3 13 6 13
6 8 16 12 17
3 4 3 8 21
1 1 3 12 42
1 0 2 1 35
0 0 3 3 53
2 1 3 10 23
0 1 3 11 4
3 0 3 8 25
1 2 6 6 44
3 2 2 4 27
2 1 8 7 39
1 1 5 7 25
0 1 7
5 46
(Oh
1
169
MATERNAL AND CHILD HEALTH CARE / RElS
50
1
40 -
30 -
H
ea
20 10 0
PPP FFS
7-
T
Hlth Dept
MD
L HMO
-r
Hosp
ER
Preference for Provider FIGURE 3. Mother’s preferred health care provider according to Medicaid status.
of sample), and hospital’s refusal to accept Medmothers were displeased with the amount of time icaid patients (83% of sample). The majority of required to go to health provider (13% versus the sample (80%)also felt it was difficult to obtain 5 % ) , information received from physician (7% Medicaid and therefore difficult to obtain health versus 2%), quality of care (8% versus 2%), and care. amount of concern expressed by nurseiphysician The mother’s preferred choice of health care (8% versus 5%). More FFS mothers than PPP provider according to her current Medicaid status mothers were dissatisfied with waiting time (24% is summarized in Figure 3. The mothers differed versus 18%). None of the differences observed in their choice of provider, with more PPP mothers between the two groups of mothers were found (41%) than FFS mothers (22%) desiring to go to to be statistically significant. Following these questions on level of satisfacthe department of health, x2(5) = 11.08, p s .05. In comparison with the other group, FFS tion, the respondents were asked if they perceived any barriers to their being able to obtain the care mothers were more interested in private physicians (36% versus 31%), hospital clinics (25% versus they felt their child needed. Significantly more 13%) and hospital emergency rooms (17% versus PPP mothers (21%) than FFS mothers ( 5 % ) in10%). In terms of past experience, 87% of the dicated that there had been a time when they PPP mothers and 70% of the FFS mothers had thought their baby needed a doctor but they could ever used CDOH services. Twenty-eight percent not get access to health care, ~ ‘ ( 2 )= 6.14, p S of the PPP group compared with 8% of the FFS .05. PPP mothers indicated that they had enmothers, stated that DOH services must be paid countered more specific barriers to care than had for, x2(3) = 7.79, p s .05. FFS mothers (21 positive responses versus 7). Being restricted to one physician by Public Aid was marginally significant, ~ ’ ( 1 ) = 3.31, p 6 .07 with Yates correction. DISCUSSION PPP mothers also were more likely to state that people on Public Aid do not receive as much good care as people not enrolled in the program, ~ ~ (=2 ) The results of the survey of inner-city mothers identified a number of differences in health status, 4.83, p S .09. In aggregate, the mothers agreed maternal perception of access to health care, and that difficulties in access to care were created with preference for health care provider according to enrollment Medicaid HMOs (68% of sample), the type of Medicaid program in which the mothers restrictions to one proyider (74% of sample), prowere enrolled. A discussion of the limitations of vider’s refusal to accept Medicaid patients (84%
170
RESEARCH IN NURSING & HEALTH
these findings is followed by suggestions of issues for further study. The generalizability of the survey results is limited by the relatively small sample size. Although the mothers interviewed were representative of the population of community women enrolled in Medicaid in terms of age, race and marital status, a larger sample would be desirable. The project team had originally planned to conduct interviews with more mothers, but were unable to meet the target sample size because of resource limitations and unanticipated difficulties in finding families. Another factor potentially limiting the validity of the survey findings is reliance on the mother as the primary source of information on health status. Ideally, the mother’s statements would have been checked against her medical record. Prepaid Medicaid providers do not, however, currently permit outside organizations to inspect their clinical documentation. In lieu of cross-checking the mother’s recall or illness and treatment with the full medical report, a check was done with the infant’s immunization record. This cross-validation documented a high level of accuracy in maternal reporting paralleling the findings of other studies of parental reports of their children’s illnesses (Bergner & Rothman, 1987). Inaccuracies in the mothers’ impressions of the accessibility and quality of their health care represents another source of bias more difficult to address than self-reports of health status. Affirmation of the respondents’ perceptions of their providers would require evidence on administrative policies and management procedures, as well as objective review of the quality of care provided. Assembling such information was beyond the scope of this project. A number of differences between the prepaid and fee-for-service experiences related to clinical and administrative practices delineate issues for future study as discussed below. From a patient care perspective, the survey findings suggest that a comparative evaluation be done of specific clinical protocols employed by providers in the various Medicaid models. Newborns enrolled in a prepaid plan had apparently contradictory experiences in that their mothers reported a lower use of newborn hospital services and primary care for infant illnesses but more \ complete and timely use of immunization services. More information is needed to determine if different models promote systematic variations in the quality of patient care. In a similar vein, objective assessments of the availability of services in the prepaid and fee-for-service Medicaid models are in order. The PPP mothers’ descriptions of difficulties in obtaining certain services may be related
to clinic management policies or may reflect some other set of factors outside the realm of the health care system. Such recommendations for future studies must be reviewed in the context of the debate on Medicaid reform and the underinsured (Wilensky, 1987). The State of Illinois Department of Public Aid has recently issued guidelines for the Medicaid program that include expanding services for pregnant women and children up to 185% of the federal poverty level (Illinois Department of Public Aid, 1989). At the same time, IDPA must find ways to contain Medicaid expenses. Exploration of clinical and administrative practices in different Medicaid models would provide information needed to strike a judicious balance between accessibility, quality, and costs for the population needing maternal and child health services.
REFERENCES Aday, L.A., Anderson, R., & Fleming, G.V. (1980). Health care in the United States: Equitable f o r whom? Beverly Hills: Sage. Alan Guttmacher Institute (1987). Blessed evenfs and the bottom line: Financing maternity care in the United States. New York: Alan Guttmacher Institute. Bergner, M . , & Rothman, M.L. (1987). Health status measures: An overview and guide for selection. Annual Review of Public Health. 8 , 191-210. Berki, S . E . , & Ashcroft. M.L.F. (1980). HMO enrollment: Who joins what and why: A review of the literature. Milbank Memorial Fund Quarterlv, 58, 588-632. Brown, E.R., & Cousineau, M.R. (1984). Effectiveness of state mandates to maintain local government health services for the poor. Journal of Health Politics, Policv and Law, 9, 223-235. Butler, J.A., Winter, W.D., Singer, J.P., & Wenger, M. (1985). Medical care use and expenditure among children and youth in the United States: Analysis of a national probability sample. Pediatrics. 7 6 , 495-507. de Vise, P. (1984). The nation’s sixteen poorest neighborhoods (Chicago regional inventory, Working Paper No. 1177). City Club of Chicago. Freund, D.A. (1986). The private delivery of Medicaid services: Lessons for administrators, providers, and policymakers. Journal ofAmbulatory Care Management, 9, 54-65. Health Care Financing Administration (1988). Medicaid and prepayment: A bibliographic essay. Washington, DC: U . S . Department of Health and Human Services, Public Health Service. Illinois Department of Public Aid (1989). A proposal for Medicaid reform: Carrying the challenge into
MATERNAL AND CHILD HEALTH CARE 1 RElS
the future. Springfield, 1L: Illinois Department of Public Aid. Kasper, J . D . (1987). The importance of type of usual source of care for children’s physician access and expenditures. Medical Care, 2 5 , 386-398. Reis, J . . & Olson, L. (1987). The Medicaid program and consumer needs: A survey among residents of a poor Chicago neighborhood. Public Health Reports, 102, 530-538. Rosenbaum, S . , &Johnson. K . (1986). Providing health care for low income children: Reconciling child health goals with child health financing
171
realities. The Milbank Quarterly, 6 4 , 442-478. U . S . Department of Health and Human Services, Centers for Disease Control (1986). New recommended schedule for active immunization of normal infants and children. Morbidity andMortality Weekly Report, 35(37), 577-579. Wallace, H . M . , Gold, E.M., & Oglesby, A.C. (1982). Maternal and child health practice: Problems, resources and methods of delivery (2nd ed.). New York: Wiley. Wilensky, G.R. (1987). Viable strategies for dealing with the underinsured. Health Affairs, 6, 33-46.
