October 1994 Volume 94, Number 4 Pediatrics

Immunization Referral Practices of Pediatricians in the United States

Holly S. Ruch-Ross, ScD* and Karen G. O'Connor

� ABSTRACT. Objective. Concerned about alarmingly low rates of immunization among some young US children as well as disturbing trends in vaccine availability and delivery, the American Academy of Pediatrics sought to examine national trends in referral for immunization among US pediatricians.

� Methods. A self-administered eight-page questionnaire on issues in childhood immunization was mailed to a random sample of the AAP's 36 000 US resident members. Four rounds of the survey yielded a sample size of 1246, for a response rate of 77%.

� Results. A majority of respondents reported referral of some patients for immunization; reported reasons for referral focused on issues of cost to the patient. Personal characteristics of pediatricians (age and gender) were not related to referral practices. Pediatricians practicing in solo or group settings were 2.8 times as likely as those in clinics or other settings to refer patients for immunization. Those in nonmetropolitan areas were nearly twice as likely as those in large metropolitan areas to refer for immunization. When setting and location were controlled, pediatricians who reported that their states had programs to provide all vaccines free or at reduced cost were much less likely to indicate that they referred patients for immunization.

� Conclusions. A majority of US pediatricians refer some of their patients to other providers for immunizations because of financial reasons. Children outside of metropolitan areas may be at particular risk for problems of availability of and access to immunizations. This study raises, once again, the issue of differential health care access for children based on payment mechanisms. Pediatrics 1994;94:508-513; immunizations, referrals, preventive care, pediatric practice.

Childhood immunization is one of the most cost effective and best accepted preventive health practices in the United States today. Approximately 98% of American children are fully immunized by school entry.[1] However, approximately 80% of the recommended vaccine doses should be received by a child by 2 years of age,[2] and, based on this criterion, very young American children are less well protected than they should be. A recent series of retrospective school record surveys in 16 states found a median of 57% of children fully immunized by age 2.[3] In some inner cities, full immunization rates for 2-year-old children are as low as 50%.[4,5] The seriousness of this situation is illustrated by the measles outbreak of 1989 to 1991. Measles had declined from nearly 27 000 cases in 1978 to about 3000 cases per year between 1981 and 1988. However, nearly 28 000 cases were reported in 1990; there were approximately 132 related deaths during the 3-year outbreak.[1]

The US Department of Health and Human Services has established a year 2000 goal of full immunization for at least 90% of children aged 2 years and under.[6] Achievement of this goal may be significantly undermined, however, by disturbing trends in vaccine availability and delivery.

Between 1989 and 1991, urban health departments reported a 22% increase in the number of children served.[4] In a 1991 survey conducted by the Children's Defense Fund,[7] community health centers across the country reported substantial barriers to immunization for poor children, particularly chronic vaccine shortages, high costs, and the decreasing adequacy of both private and public insurance coverage. Pediatricians and family practitioners in Dallas County, Texas reported that they had increased referrals for immunizations between 1979 and 1988, most frequently to public health clinics.[8] The most commonly cited reason for referral was patient inability to pay. This primary focus on cost to patients as the reason for referral was also reflected in a recent survey of pediatricians in Washington State.[9]

Those who provide health care to children are concerned that referral for immunization interferes with continuity of care.[10] Moreover, the likelihood that a child will be adequately immunized is substantially improved if the child has a single primary care provider who is able to monitor immunization status and provide immunizations as needed.[2] The possibility that referral reduces the likelihood that a child will receive other preventive care services, including subsequent immunizations, is very troubling. In order to examine national trends in referral for immunization among pediatricians, the American Academy of Pediatrics (AAP) undertook a survey of its membership in the spring of 1992.

METHODS

A Periodic Survey of the membership of the AAP is conducted three or four times annually on topics of interest to the AAP and its members. The survey is administered to a random sample of approximately 1000 of the AAP's membership of 36 000 US residents, which includes approximately 75% of US board-certified pediatricians; those selected to participate in a Periodic Survey are not selected again for a period of at least 4 years.

The present study was the 18th Periodic Survey, fielded during the spring of 1992. A sample of 1600 was drawn to ensure an adequate number of responses from pediatricians in direct patient care. An original and three follow-up mailings, each including a postage-paid return envelope, resulted in a total of 1246 completed questionnaires for a final response rate of about 77%.

Periodic Survey #18 employed a self-administered eight-page questionnaire. The survey included questions regarding missed opportunities for immunization as well as the immunization referral questions included in the present analysis.

Factors examined in this analysis include age, gender, practice location, practice setting, and the availability of a state vaccine program. Practice location is divided into three levels: metropolitan area of 1 million or more population, metropolitan area with a population between 50 000 and 1 million, and nonmetropolitan areas with population less than 50 000. Practice (primary employment) setting also has three categories: solo or two-person practice, group practice or HMO, and clinics and other settings. (In the present data set, HMO and group practice settings cannot be separated out. In previous surveys, pediatricians in staff model HMO settings comprised approximately 6% of the sample.) Referral practices in this analysis include current referral of patients for immunization, reported change in referrals in the past 10 years, referral of specific patient groups, and reported reasons for referrals.

Pediatricians who reported that they do not provide preventive care services, about 25% of the sample, were excluded from the analysis. These pediatricians are exclusively in subspecialty practice, or are primarily engaged in teaching, administration, or research. Pediatricians in training (Residents) were also excluded from the analysis, primarily because their concentration in particular locations and settings may skew the findings.

Bivariate relationships were tested using contingency table analysis (chi square) or comparison of means (t test) as appropriate. Logistic regression was used to examine multivariate relationships between the independent variables and referral for immunization coded as a dichotomous variable.

RESULTS

Characteristics of Respondents

The characteristics of the respondents (Table 1) are, in general, similar to those of respondents to previous periodic surveys. Their mean age is about 42 years. Nearly 39% of respondents are female, and a substantial minority (42%) practice in large metropolitan areas with populations in excess of 1 million. A majority (52%) reported a solo or group practice as a primary employment setting. These characteristics are not independent; female Fellows of the AAP (FAAPs) tend to be younger, and are more likely to practice in clinic settings and to be in general practice than their male counterparts.

State Vaccine Programs

At the time of the study there were 16 states that had either universal or Medicaid-only programs in place to provide all or some of the recommended vaccines (diphtheria, tetanus, pertussis vaccine; oral poliovirus vaccine; measles, mumps, rubella vaccine; Haemophilus influenzae type b vaccine) to physicians free of charge or at a reduced cost; in five other states some form of a vaccine program was scheduled to be enacted during the year of the survey. Just over one-third (34.7%) of respondents reported that the states in which they practice provide some or all vaccines; more than one-third (37.6%) did not know whether their states had free vaccine programs (Table 1). However, respondents were not particularly accurate in reporting the existence of a state program. About 53% of pediatricians living in states which do, in fact, provide at least some vaccines reported that their states provide some or all vaccines; 13.5% of pediatricians in these states reported that their states do not have such programs. In states which do not provide vaccines, approximately one-quarter (26.1%) of respondents reported that their states do so (not shown).

Immunization Referral Practice

A total of 707 practicing pediatricians in the sample reported that they provide preventive care. Nearly half (44.8%) of these respondents reported that they do not refer any patients for immunizations; only 3.3% refer all patients (Table 2). Although pediatric subspecialists were not more likely than general pediatricians to report that they refer some or all of their patients for immunization, they were most likely to indicate that they refer all patients for immunizations (not shown).

Referral practices did not vary by personal characteristics of the pediatricians, but they varied significantly by practice characteristics. A much larger proportion of pediatricians practicing in large metropolitan areas (53%) reported that they do not refer any patients than those in small metropolitan areas (44%) or nonmetropolitan areas (31%). When pediatricians in private settings are separated from those in clinics and other settings, the proportion reporting that some or all children are referred for immunization is substantially higher than the overall figure; about 67% of those in solo or two-person practices and 61% of those in group practices or HMOs report referring children. As would be expected, pediatricians practicing in clinics and other settings, which include many public settings to which children would be referred for immunization, are substantially less likely to report referring any children for immunization (32%). Awareness of a state vaccine program was also significantly associated with referral of patients for immunization; only 30% of respondents who reported a state vaccine program which provides all vaccines referred children for immunization. In contrast, 75% of those who reported that their states provide no vaccines referred children for immunization.

Pediatricians who provide preventive care were asked what percent of patients who required immunizations were referred to another source in the last month. More than half (60%) reported that 10% or fewer patients were referred; only 9% referred more than half of their patients for immunizations (not shown). The reported mean percentage of patients referred by pediatricians who referred all or some patients was 20%.

Approximately 43% of responding pediatricians indicated that referrals for immunization have increased in the last 10 years; very few (about 7%) reported a decrease in referrals during this period (Table 3). More pediatricians outside of large metropolitan areas reported that referrals have increased in the last 10 years (34%, 44%, and 49% for large, small, and nonmetro, respectively). Very substantial proportions of both groups of pediatricians in private settings report that referrals have increased in the last 10 years (48% for solo or two-person and 43% for group or HMO practices). Most respondents felt that referral for immunization had not affected the number of preventive care visits in their practices, although about 18% reported that the number of such visits had decreased as a result of referral for immunization (not shown). Those in nonmetropolitan areas (30% vs 20% small metro and 12% large metro, P < .05; not shown) were most likely to report that preventive care visits have decreased as a result of referring patients to another source for immunization.

Not all patients are equally likely to be referred (Table 4). Among pediatricians who see children in each of the following groups in their practices, about 87% refer some or all uninsured children, about 70% refer migrant children, about 61% refer Medicaid recipients, and about 57% refer some or all insured children. Most pediatricians who refer some or all patients for immunization make referrals to the Public Health Department clinics (95%; not shown). Among pediatricians who serve Medicaid patients and who refer some or all patients for immunization, those in metropolitan areas (67% and 64% in large and small metros) were much more likely to refer these patients for immunization than those practicing in nonmetropolitan areas (47%). Only referral of uninsured children varied by practice setting. About 93% of those in group practice or HMO settings and 85% of those in solo or two-person practices who see uninsured children report referring them for immunization. However, more than half (54% to 56%) of these private practitioners report referring even insured children for immunization some of the time.

The most important reasons for referral for immunization, by a substantial margin, were those related to the cost to the patient (Table 5). Well over half of respondents indicated that patient refusal due to cost (56%) or request for referral due to cost (65%) were very important reasons for referral. By contrast, cost to the pediatrician (28%), unavailability of vaccine (10%), and concern about liability (2%) were much less frequently rated very important by the pediatricians responding to this survey. Small metropolitan and nonmetropolitan pediatricians were more likely than their large city counterparts to indicate that patient refusal due to cost (45%, 60%, and 63%, for large, small, and nonmetropolitan areas, P < .05; not shown) and request for referral due to cost (56%, 71%, and 65%, P < .05; not shown), as well as the high cost of vaccines to the physician (13%, 32%, and 36%, P < .001; not shown) were very important reasons for referral for immunization.

Logistic regression results were very similar to the bivariate findings regarding referral for immunization. Personal characteristics of pediatricians were not predictive of referral, nor did their presence in the model affect the relationships between practice characteristics and referral. Table 6 presents the model for referral for immunization by pediatricians who provide preventive care. Controlling for location and awareness of a state vaccine program, pediatricians in solo or group settings were almost three times as likely to report referring for immunization as those in clinic and other settings. Pediatricians in nonmetropolitan areas were nearly twice as likely as those in large metropolitan areas to report referring some or all of their patients, even when primary setting and awareness of a state program were controlled for. Although all other groups were substantially more likely to report referring for immunizations than those who indicated that their states provided all vaccines, this difference was smallest for those who did not know whether their states provided vaccines.

DISCUSSION

A substantial proportion of pediatricians who are members of the AAP report that they refer some children to another provider for immunization. The most frequently reported place to which children are referred is a public health clinic. Immunization and referral practices of pediatricians vary significantly by location of practice and employment setting. Pediatricians outside of large metropolitan areas are most likely to refer patients and refer higher proportions of patients, and more of these respondents indicate an increase in referrals in the last 10 years. This suggests that issues of availability of and access to immunizations is a growing problem, particularly for children who live in nonmetropolitan areas.

This survey revealed a lack of knowledge regarding state vaccine programs on the part of pediatricians who provide preventive care. Perhaps this result is not so surprising, given the limited effort to publicize vaccine programs in some states. There is a clear need for better dissemination of information about these programs to pediatricians. This is particularly true in light of the finding that pediatricians who were aware of a state program to provide all vaccines were much less likely to refer patients for immunizations than all other groups.

In this survey, as in previous studies,[8,9] cost to the patient appeared to be the most important reason for referral for immunization. Cost of completion of the recommended immunization series in the private sector is estimated at $324.[11] For the half of US children who receive their immunizations in the public sector, the cost is approximately $122; this cost is not borne by the patient.[11] Although immunizations are clearly cost effective in the aggregate, the impetus to refer individual patients who would otherwise pay out-of-pocket, particularly those of limited means, is clear.

Pediatricians outside of large urban areas were more likely to rate cost to the patient and the physician as a reason for referral for immunization than their urban counterparts. More nonmetropolitan pediatricians also perceived a deleterious effect on preventive care visits when children are referred elsewhere for immunization. Once again, it appears that there may be important geographic differences in trends in childhood immunization which require more focus on access for children who do not live in major population centers.

As might be expected, pediatricians in private settings were much more likely to report referring some or all patients than those practicing in other settings. However, the proportion of these pediatricians reporting an increase in referrals in the last 10 years approached one half, suggesting that more and more families are experiencing the cost of immunization as a financial hardship. This finding is particularly noteworthy, given that HMO pediatricians, who would be expected to refer very few children, are included with those in other group practice settings.

The finding that a majority of respondents report referring privately insured patients reflects the reality that many health insurance plans still do not cover the costs of preventive care services, including immunizations. A somewhat larger proportion of pediatricians in this sample reported referring Medicaid recipients for immunizations, but an overwhelming majority of those who see uninsured children, in every location and every setting, report referring these children elsewhere for immunizations. Thus, this study raises, once again, the issue of differential health care access for children based on payment mechanisms.

REFERENCES

----------------

(C) 1996 American Academy of Pediatrics

---------------------------------------------------------------------------------------------------

------------------------------------------------------------------------------------------------------------