Objective. Pediatric residency training programs need regular revision, but there is little evidence to support changes. The recent reduction in time spent in the newborn nursery may negatively affect the care of infants in pediatric practice. This survey was undertaken to assess the preparation of recent graduates of training programs with respect to newborn care in their current practice.
Methods. A nationwide survey was sent to 1997 pediatric residency graduates in May 1999. The American Medical Association Masterfile was used to identify pediatricians who met the criteria. The survey included questions about preparation and confidence in topics obtained from the Residency Review Committee guidelines for the newborn nursery. There were questions about duration of nursery experience, learning barriers, teaching preferences, current nursery responsibility, and practice location along with other information about the respondents.
Results. After 2 mailings, 801 surveys were received. Practice distribution was comparable to those who took the pediatric boards for the first time in 1997, and response was nationwide. Overall, 45% attend deliveries and 71% care for sick newborns. Respondents with only 1 month of normal newborn nursery felt this was insufficient preparation. Bedside teaching by faculty members was rated most highly. The most commonly rated areas of poor preparation were prenatal visits, perinatal infections, and management of breastfeeding issues.
Conclusions. The current structure of nursery training in many residency programs may be insufficient preparation for primary care practice. Results of this and future surveys of recent graduates provide evidence to direct changes in residency training.
The Future of Pediatric Education (FOPE) II project of the American Academy of Pediatrics has renewed interest in how we train residents to practice pediatrics. The initial FOPE statement published in 1978 recommended an increase in time spent in the ambulatory care setting and a decrease in subspecialty training.1 A survey of residency program directors in 1983 showed slow response to the recommendations, with few revisions in training programs.2 A recent review article on pediatric residency education looked at the opinion of pediatric program directors on the impact of the 1997 Residency Review Committee (RRC) guidelines.3 Among other things, the directors called for more outcome measures than simply the assurance that residents meet time requirements in specific fields. It was also reported that they would like more flexibility in RRC guidelines to tailor their programs in terms of regional workforce issues. The article calls for more outcome studies on a variety of educational issues so evidence-based changes can be made in programs. Although the input of program directors is essential, most of these people are practicing in an academic setting. When considering changes to residency program structure it is important to look at the experience of recent graduates who are now in primary care practice.
A survey of graduates is an ideal but not novel way to determine the effectiveness of training programs and target areas for revision. There have been several surveys of individual training programs, most of which involved graduates, that span many years of training with results difficult to generalize to other programs.4–9 Neonatology training was highly regarded in most of these surveys, but the quality of normal newborn nursery (NNN) training was not assessed. Infants are the foundation on which a general pediatric practice is built, with more well child visits occurring in the first 2 years of life then the entire next decade. The move to more outpatient pediatric training has resulted in a decrease in the number of inpatient rotations, including time in the neonatal intensive care unit (NICU). The concern is that there is not enough time allotted to neonatal care in the current structure of residency training programs. This is the first nationwide survey of recent graduates of residency training programs to specifically target newborn issues. The study was undertaken to determine how prepared the graduates of pediatric residency felt to care for newborns in their primary care practice.
A nationwide survey of newborn medicine training in residency was conducted for pediatricians in practice 2 years at the time of the initial mailing. Recipients were identified by the American Medical Association Masterfile as 1994 medical school graduates entering pediatric residency programs. The 2-page survey was formatted after the American Academy of Pediatrics Periodic Survey number 40 (distributed in 1998) and took less than 15 minutes to complete. The first page of the survey included questions about preparation and confidence in topics obtained from the Residency Review Committee (RRC) guidelines for the newborn nursery. The second page included questions that address learning barriers and teaching preferences in the nursery along with information on practice setting, location, sex, age, and place of training (optional). The survey was approved by the Children’s National Medical Center Institutional Review Board and piloted in graduates of that program. The first mailing began in May 1999, and the second mailing was completed in June 2000. An attempt was made to locate pediatricians with erroneous addresses on file by search of the American Academy of Pediatrics Fellows directory. The survey was sent a second time to nonresponders. The return envelopes were numbered for tracking purposes, with the survey immediately separated from the envelope on receipt to maintain confidentiality. The zip code of the primary practice site designated on the completed survey was entered into a database of 1990 census information on the US Gazetteer Web site (http://www.census.gov/cgi-bin/gazetteer) to distinguish rural from urban locations. A chi-square analysis was done to compare results between groups. A copy of the survey instrument is included in the Appendix.
A total of 801 of the initial 1690 distributed surveys were received after 2 mailings; 268 returned as undeliverable, for a final response rate of 56%. Survey respondents were from all areas of the United States and represented 151 pediatric residency training programs. Male pediatricians compose 34% of the total, and 66% are female. A primary care practice setting was identified by 76% of the respondents (Fig 1), with an overall practice distribution comparable to those who took the pediatric boards for the first time in 1997.10
Perception of Duration of Newborn Training (Fig 2)
Respondents were asked to recall the number of months of training that occurred in the NNN and NICU. The questions states, “How many required months of well-infant nursery did you complete during residency?” The respondents were to determine what constituted a month of training in their program, with no specific definition provided, and the same question was asked for the duration of NICU training. More than 70% of the primary care respondents (N = 559) spent either 1 or 2 months in the NNN during pediatric training. In addition, respondents were asked to provide an opinion of whether their nursery time was too little, the right amount, or too much. The majority (67%) of those with 1 month of NNN felt it to be insufficient preparation for primary care practice (Fig 2). In contrast, 72% and 75% of those with 3 and 4 months of NNN, respectively, felt it to be the right amount of training. Responses for number of months of NNN training ranged from 1 to 10 months. These numbers probably reflect both on-service and cross-cover time for many respondents.
The number of months spent in the NICU during training ranged from 2 to 12, with 27% of primary care respondents reporting 6 or more months of NICU time. Again, those with the greatest number of months were likely to have included cross-cover time. The largest number of respondents (n = 175, or 33%) had 4 months of NICU training, 60% of whom felt it to be the right amount. The majority of those with 5 months or more of training felt it to be too much (Fig 3).
Many of the surveys included comments from the respondents. A frequent remark regarded the recent restructure of their programs to provide less NICU time. Several voiced concern that they would be at a disadvantage without the additional NICU experience.
Preferred Teaching Methods in the Nursery (Fig 4)
Bedside teaching by faculty was rated the highest of teaching methods in nursery, with 69% finding it very helpful and only 2% not helpful. Bedside teaching by residents or fellows was rated as very helpful by 38% and 35%, respectively. Rounds were the next highest teaching method, with 53% finding them very helpful and 42% helpful. Lectures were rated as helpful by 57% and very helpful by 34%.
Preparation and Confidence on RRC Guideline Topics
Primary care pediatricians most commonly rated the following topics as areas of poor preparation during training: prenatal visits (43%); perinatal infections, including syphilis (17%) and human immunodeficiency virus (22%); breastfeeding issues (21%); and preterm infant follow-up (17%). Of the male primary care pediatricians, 75% felt not or only somewhat confident managing breastfeeding issues, compared with 64% of the female primary care pediatricians (not significant, P > .1). Topics for which over 80% of respondents indicated very good preparation include hyperbilirubinemia (87%) and sepsis risk factors (86%). Overall, differences between primary care respondents and total respondents did not differ with respect to opinions on preparation and confidence. More than half of all primary care respondents felt not or only somewhat confident with nutritional issues such as formula intolerance (54%) and preterm formula use (55%).
Delivery Room and Resuscitation Experience
Overall, 45% of primary care pediatricians attend deliveries in their current practice, and 71% care for sick infants in the nursery. Only 35% of those who felt they had too little NICU training attend deliveries in their current practice, compared with 52% of those who felt they had the right amount of NICU training (P < .001). Only a small percentage of primary care doctors felt they had too much NICU training, but even more of these physicians (55%) attend deliveries. The percentage of pediatricians who reported a mock code experience in training was 63%, with 89% finding them helpful or very helpful. Fifty-six percent of those with mock code experience attend deliveries in their current practice. However, there was no difference in the confidence with delivery room resuscitation procedures between those with mock code experience and those without.
Many factors must be considered when a residency program is planned. Training must be diverse and appropriate to the future practice of trainees. In this study we examined the opinion of recent graduates of the impact of residency training on the care of newborns in their current practice.
The amount of training in the NNN or NICU varied widely among respondents. This data are complicated by the diversity of training programs with respect to the division of time spent on rotations (ie, some programs use calendar months, some use “units” or “blocks,” which differ in the total amount of days). Respondents were left to determine what constituted a month of training in each specialty, and it is clear that cross-cover time falsely elevated the total time designated. The reported amount of NNN training in programs was insufficient according to many respondents, with the ideal amount of time closer to 3 months than the commonly allotted 1 month, usually in the first year of training. It may be helpful to consider a block of NNN during the later years of residency, when pediatricians have more confidence and can understand the impact of this time on their future practice. Time spent in the NICU is not a substitute for time in the NNN, where the skills necessary for the care of healthy newborns and counseling of new parents are gained. Inclusion of topics pertinent to the NNN in the core conference sessions of residency training programs would help expand training in this area but cannot completely substitute for direct patient contact.
Responses show that some programs far exceed the current RRC guidelines for NICU time (limited to 6 months of ICU time including 2 months of pediatric intensive care). This data are consistent with a survey of chief residents in 1997 that revealed many programs are not compliant with recommendations for the amount of NICU training.11 Many primary care respondents voiced concern that their programs had recently limited the amount of NICU training to a level that would be dangerously low for those entering primary care practice. Some commented that although they did not particularly enjoy their NICU experience, it was essential for the care of infants in their current practice and should not be reduced. The NICU experience in many programs must compensate for the lack of NNN time with respect to learning to manage common problems encountered in newborns. However, it often excludes the opportunity to be the initial diagnostician when these problems arise because many term infants arrive in the NICU with a problem already identified in the NNN by another doctor. An improved balance must be achieved in training programs, providing adequate amounts of both NNN and NICU experience.
A much larger number of primary care pediatricians attend deliveries and care for sick infants than expected. Much of the experience dealing with the care of sick infants is gained during NICU rotations. Resuscitation skills are more likely to be learned during NICU rotations because of the more frequent occurrence of code situations in that setting than in the delivery room. The NICU is a more amenable environment in which to practice these essential skills as well. Those who felt their NICU experience comprised the right amount or too much time were more likely to attend deliveries in their current practice. The high number of pediatricians attending deliveries in this survey may be unique to the early years of practice but indicates the importance of strong preparation for delivery room resuscitation. This information is consistent with that reported in 1996 from a survey of Texas pediatricians that found that a large percentage of both rural and nonrural pediatricians cared for newborns who needed level II care.12 The high number of pediatricians caring for sick infants reflects the changing dynamics of health care and the importance of pediatric residency training that addresses the care of sick and well newborns.
NICU training is unique in that it helps residents learn the care of preterm infants and understand concerns that arise with their long-term management. The earlier discharge of these infants in recent years necessitates frequent follow-up with a pediatrician in the first months of life. These physicians need knowledge of problems unique to growing premature infants including developmental changes in the pulmonary, neurologic, and immune systems as well as nutritional management. This is a valuable aspect of NICU training and warrants continued emphasis during residency.
Overall, most of the respondents felt they were well prepared with respect to the specific areas of newborn training recommended by the RRC. Areas that warrant improved preparation include prenatal visits, management of breastfeeding problems, preterm follow-up, and infant formula changes. Confidence varied on many topics, but particularly concerning is the lack of confidence in managing breastfeeding problems. Mothers commonly have difficulty with initiating breastfeeding, and many hospitalizations in the first weeks of life result from breastfeeding problems.13 Identifying and treating lactation problems must be a primary focus during NNN training and should continue to be offered as a topic for continuing medical education for practicing primary care pediatricians.
The quality of teaching in the nursery affects the overall educational experience. The teaching methods favored by the respondents were the presence of faculty at the bedside and during rounds. It is important to devote uninterrupted time each day to teach the house staff during on-service months, which are often filled with other clinical, administrative, and research responsibilities. Another teaching barrier at our institution is the current schedule of the house staff that takes them away from the units several days a week for other responsibilities. We needed to be creative and find alternatives to scheduled afternoon lectures to reach the residents. Our new methods include finding teachable moments during rounds and designating a period of time after rounds that is spent at the bedside of an interesting patient to expand the discussion of the disease process, physical findings, and treatment.
The limitation of this study is that the findings are based only on the experience and opinion of survey respondents. Although the response rate was only 56%, it was nationwide and included pediatricians with demographics equal to those who took the pediatric boards for the first time in 1997. The nonresponders therefore are likely to be similar, but that cannot be definitively ascertained. The actual response rate may have been higher if the number of surveys sent to incorrect mailing addresses could have been accurately determined, which was impossible because of local post office practices. The pediatricians in this study would not have undergone training influenced by the latest RRC guidelines, but responses are important to the 1997 recommendations and should be considered as training programs are revised.
It is essential to look at the experience of trainees and assess how well we are preparing them for future practice. As we strive to use evidence-based medicine as a guide for clinical management, we should hold these same standards to the evaluation and development of training programs. Increased focus should be placed on NNN training, with attention to management of breastfeeding problems, identification of common problems, and anticipatory guidance for the parents of newborns to provide new pediatricians with an adequate level of preparation and confidence. Emphasis should remain on NICU training to provide skills necessary for newborn resuscitation and management of the complications of prematurity and other conditions that occur in the newborn period after NICU discharge. Overall, our preparation of these future pediatricians may be inadequate for the demands of infant care in primary pediatric practice.
This work was supported in part by a grant from Mead Johnson Nutritionals, which provided funding for survey duplication and mailing.
We would like to acknowledge all those who provided advice and support for the design of this project and review of the survey instrument and manuscript, including Larrie Greenberg, MD; Tina Cheng, MD; Jill Josephs, MD; Bernhard Wiedermann, MD; and Billie Lou Short, MD.
Lastly, we would like to thank the pediatricians who participated in this survey and provided helpful insight and advice.
- Received May 24, 2001.
- Accepted May 13, 2002.
- Reprint requests to (D.M.W.) Department of Neonatology, Holy Cross Hospital, 1500 Forest Glen Rd, Silver Spring, MD 20910. E-mail:
- ↵Task Force on Pediatric Education. The Future of Pediatric Education. Evanston, IL: American Academy of Pediatrics; 1978
- ↵Weinberger HL, Oski FA. A survey of pediatric resident training programs 5 years after the Task Force report. Pediatrics.1984;74 :523– 526
- ↵Mulvey HJ, Ogle-Jewett EAB, Cheng TL, Johnson RL. Pediatric residency education. Pediatrics.2000;106 ;323– 329
- DeAngelis C, Duggan A, Oski F. Twenty-five years of residents: what and where are they now. Pediatrics.1990;85 :10– 16
- Wender EH, Bijur PE, Boyce WT. Pediatric residency training: ten years after the Task Force report. Pediatrics.1992;90 :876– 880
- ↵Haggerty RJ, Janeway CA. Evaluation of a pediatric house officer program. Pediatrics.1960;26 :858– 861
- ↵Tunnessen WW. Workforce: results of 1998–99 tracking. Pediatric Diplomates: A Newsletter From the American Board of Pediatrics. Chapel Hill, NC: American Board of Pediatrics; 1999:5
- ↵Feigin RD, Drutz JE, Smith EO, Collins CR. Practice variations by population: training significance. Pediatrics.1996;98 :186– 190
- Copyright © 2002 by the American Academy of Pediatrics