 |
|
Deliveries by Family Physicians in Connecticut
Results of a Practice-Based Research Network
Keith Sinusas, MD
Arch Fam Med. 2000;9:434-438.
ABSTRACT
| |
Objective To obtain descriptive data on deliveries performed by family physicians in Connecticut during 1 year.
Design All family physicians in Connecticut who provided obstetric care were personally contacted to enlist their participation in a research network. All such physicians agreed to participate and were instructed to mail a 15-item delivery data card to the author following each delivery that they participated in during calendar year 1996.
Main Outcome Measures Method of delivery, use of obstetric intervention (eg, oxytocin administration, episiotomy, and epidural anesthesia), and obstetric consultation rates.
Results Thirty-two physicians, representing 9.0% of the family practice workforce in Connecticut, provided obstetric care. These 32 physicians, who practice in only 7 of the 31 acute care hospitals in the state, delivered 478 neonates during 1996. Most deliveries were by the spontaneous vaginal route, with forceps and vacuum used in 2.1% and 5.0% of vaginal deliveries, respectively. The primary cesarean section rate of these family physicians was low at 5.6% compared with a statewide rate for all providers (physicians and certified nurse midwives) of 12.4% (odds ratio, 0.42; 95% confidence interval, 0.28-0.63). Rates of certain obstetric interventions, such as oxytocin administration, epidural anesthesia, and episiotomy, varied greatly by hospital. Obstetric consultation rates also varied greatly among hospitals, ranging from 12.8% to 49.6%.
Conclusions Family physician involvement in maternity care in Connecticut is low. This study confirms a low rate of instrument-assisted and cesarean delivery in births attended by family physicians. Use of obstetric interventions and obstetric consultation varies greatly within the state.
INTRODUCTION
OBSTETRIC CARE is an important part of the care provided by family physicians. Previous studies have looked at the style of practice of family physicians who provide maternity care. Some of these studies have been set in a single practice,1 a single rural hospital,2 a single teaching hospital,3 a prepaid health maintenance organization,4 a cross section of hospitals in various states,5 and the military.6 To date, no study has looked at the delivery room activities of all family physicians in a single state. This report is a descriptive study of the deliveries performed by family physicians in Connecticut during 1996. It is the result of the cooperative efforts of a research network, which included all family physicians in Connecticut who provide maternity care to their patients.
METHODS
The maternity units of all acute care hospitals in Connecticut were contacted to determine if there were family physicians on staff who had privileges in obstetrics. Each of the physicians identified was then personally contacted to see if he or she would be willing to participate in a research network that sought to gather data on their deliveries for the 1996 calendar year. All agreed to participate.
A 15-item delivery data card was prepared, which could be completed by each physician at the end of a delivery. Data were compiled only on liveborn neonates delivered during the study period. The variables included basic demographics about the patient, identification of the delivery type, use of technology such as electronic fetal monitoring, use of interventions such as epidural anesthesia, oxytocin, and episiotomy, and use of consultation from obstetric colleagues. Apgar scores were included as a limited way to obtain data on fetal outcomes.
The delivery data card was designed so that it could be folded over and taped closed for mailing. The cards were stamped and addressed for convenient mailing by each participating physician. Once the cards were received, subsequent data entry and analysis were performed. It became clear that the bulk of the deliveries by family physicians in Connecticut were performed in 3 hospitals. Accordingly, a careful delivery log review was performed at each of these institutions, and several new cases were discovered. These were all included in the final statistics.
All data were entered into a database program (Microsoft Access; Microsoft, Redmond, Wash) and then analyzed. Odds ratios for cesarean section rates were computed using a public domain statistical software package (Epi Info, Centers for Disease Control and Prevention, Atlanta, Ga).
Since most deliveries by family physicians in Connecticut are performed at 3 institutions, the data were analyzed as a statewide aggregate and in 3 subsets (hospitals A, B, and C). Hospital A is a community hospital with a university-affiliated family practice residency, and hospital B is a community hospital with a university-administered family practice residency. Delivery data at both these hospitals reflect maternity care provided by private family physicians and family practice resident deliveries that are directly supervised by the family physician faculty of each institution. If a resident performed a delivery, it was ascribed to the attending family physician in attendance at the delivery. Hospital C is a rural community hospital that has no trainees, and the deliveries reported are all by practicing family physicians. The remainder of the family physician deliveries in Connecticut are not considered separately but are included in the statewide aggregate figures.
RESULTS
There are 31 acute care hospitals in Connecticut. Seven of the 31 hospitals had family physicians on staff who delivered babies during calendar year 1996. Thirty-two family physicians who provided this service to their patients were identified at these institutions. This represents 9.0% of the 354 family physicians known to practice in the state (Connecticut Academy of Family Physicians, Bloomfield, 1996 Membership Database). Family practice residents are not included in this figure of 32 physicians. Any delivery by a family practice resident is ascribed to the attending family physician present at the delivery. Family practice resident deliveries supervised by obstetricians are not included.
Data delivery cards were completed on 478 deliveries by this group of 32 family physicians during the study period. The number of deliveries performed per family physician during this 1-year period ranged from 1 to 48, with a mean of 13.1 and a median of 13 deliveries per year.
The demographics of the patient population are detailed in Table 1. The mean age of the patients was 25.8 years, which did not vary much among the 3 principal hospitals. The racial mix, however, did vary, with hospital C, located in a rural setting, serving a primarily white population (98.7%) and hospital B, an inner-city institution, serving a primarily nonwhite population (37.6% Hispanic and 21.4% African American). Most women were multiparous, but this too varied somewhat among the hospitals (Table 1).
|
|
|
|
Table 1. Patient Population
|
|
|
A total of 442 (92.5%) of the 478 deliveries reported by the participating family physicians were vaginal. The primary cesarean delivery rate of the study population was 5.6%, with rates in hospitals A, B, and C being 7.6%, 6.0%, and 1.9%, respectively. There were very few instrument-assisted deliveries reported, with 24 patients (5.0%) being delivered of neonates by vacuum assist and 10 patients (2.1%) being delivered of neonates with the aid of forceps. Further details on the mode of delivery can be found in Table 2.
The use of various obstetric technologies and interventions was examined in this study. Electronic fetal monitoring was used in 461 (96.4%) of the 478 labors and was consistently high in all 3 principal institutions. Epidural anesthesia was used for pain management in approximately one fifth of the labors reported. There was considerable variation among hospitals, with reported rates in hospitals A, B, and C of 18.5%, 44.9%, and 3.4%, respectively. Oxytocin was used to either induce or augment labor in 201 (42.0%) of 478 women, with hospital A reporting the lowest use rate at 29.2%. Episiotomy was performed on 196 (44.3%) of 442 patients who delivered vaginally. Surprisingly, one institution, hospital C, had a very high rate of episiotomy use, with more than two thirds of the women undergoing this procedure during delivery. Further detail on the use of these technologies and procedures can be found in Table 3.
|
|
|
|
Table 3. Interventions and Procedures
|
|
|
Consultation from a specialist in obstetrics and gynecology was obtained intrapartum in more than one third of labors (36.2%). This varied from a low consultation rate of 12.8% in hospital C, the rural hospital, to a high of 49.6% in hospital B, the inner-city facility. Mandatory consultations are required in certain institutions. For instance, in hospital A, obstetric consultation is required for use of oxytocin, and therefore all 50 women who had oxytocin administered during labor at that hospital had an obstetrician involved in their care. On the other hand, in hospital C, where family physicians are not restricted in their use of oxytocin, only 11 (13.9%) of 79 women had an obstetric consultation while receiving oxytocin. Table 4 shows the data on consultations in detail.
|
|
|
|
Table 4. Obstetric Consultations
|
|
|
Neonatal outcomes were not examined in detail, except to report the Apgar scores at 1 and 5 minutes after delivery. The only perinatal death occurred in a premature labor at 21 weeks' gestation. Apgar scores of 6 or less at 5 minutes occurred in 7 (1.5%) of 478 liveborn neonates, and scores of 7 or less occurred in 13 (2.7%) of 478. Scores greater than or equal to 8 at 5 minutes, generally considered a favorable outcome, were noted in 97.3% of liveborn neonates.
COMMENT
Family physician involvement in maternity care in Connecticut is low, with only 9.0% of active practitioners including it in their practice. This is well below the national figure of 25.2%, according to 1996 data available from the American Academy of Family Physicians (Hospital Privileges Survey, May 1996). Possible explanations for this low number may be similar to those quoted in previous studies, such as lifestyle issues, malpractice premiums, and competition with or restriction of privileges by obstetrician-gynecologists.7 In Connecticut it is possible that the relationship with obstetrician-gynecologists is a factor. Being a small state, there is ready access to obstetrician-gynecologists in every county, with practitioners on the staff of every hospital. This is very unlike some Midwestern and Western states, which may have no obstetrician-gynecologists for many miles. Despite this potential barrier of competition, family practice maternity care remains viable, with 1 of our participating physicians having attended 48 births during 1996.
The cesarean section rate among family physicians has been reportedly low in previous studies. In 1995, Hueston8 reported a nonelective cesarean section rate of 9% among family physicians compared with 16% among obstetricians in the same institutions. The present study confirms a low rate among Connecticut family physicians, with a primary cesarean section rate of 5.6%. When repeat cesarean sections are included (most of those from failed vaginal births after cesareans), the Connecticut family physician total is only 7.5%. Statewide figures in Connecticut are available for fiscal year 1996 (October 1, 1995, through September 30, 1996). The state cesarean birth rate for all providers was 19.5%, with a primary cesarean section rate of 12.4%.9 Statistical analysis of the primary cesarean birth rate of family physicians vs all providers reveals an odds ratio of 0.42 (95% confidence interval, 0.28-0.63).
The number of vacuum-assisted and forceps deliveries was very low in this study, only 7.1% of vaginal deliveries. Wanderer and Suyehira4 reported use of forceps or vacuum by family physicians in 30 (16.8%) of 179 deliveries. National data from 1987 show that 15.5% of women had instrument-assisted deliveries.10
Electronic fetal monitoring use was high in this population (96.4%), despite evidence in large randomized clinical trials showing no benefit to its use in low-risk patients.11-12 Decreased electronic fetal monitor use has been championed by Smith et al13 in their important article on obstetric care, but regretfully it has been hard for physicians to buck the tide of technology in this area.
There was considerable variability in the state with regard to the use of certain interventions and procedures. Use of oxytocin for augmentation ranged from 12.9% in hospital A to 29.9% in hospital B. This may be explained by the restrictive privileges in hospital A, where all family physicians are required to have an obstetric consultation before administering oxytocin to their patients. Perhaps the family physicians in hospital A held off on the use of oxytocin for their patients in an attempt to avoid what they believed were unnecessary consultations.
Epidural anesthesia use was similarly disparate among institutions, with family physicians in hospital B using it in 44.9% of patients and those in hospital C using it in only 3.4%. More detailed analysis of hospital C shows that, although it had the lowest rate of epidural anesthesia use, it also had the highest rate of oxytocin use. This interesting combination of aggressive use of oxytocin and limited use of epidural anesthesia in hospital C may account for their low rate of primary cesarean birth (1.9%). This would be an interesting combination to look for in other institutions around the country to see if this pattern is reproduced.
The use of episiotomy in Connecticut also varies widely, ranging from 26.8% in hospital A to 69.8% in hospital C. In the United States, episiotomy is used in more than half of all vaginal births,10 whereas in Canada a lower overall rate of 37.7% is found.14 A future goal of the research network might be to determine if educational interventions could change the behavior of physicians in hospital C with regard to episiotomy.
A major function of all family physicians is the coordination of care, which includes the appropriate use of consultants.15 In this study, obstetric consultation rates varied from 12.8% in hospital C to 49.6% in hospital B. The family practice literature likewise shows variation in consultation rates. Hueston16 reported a combined consultation rate of 15.4% in 5 different medical centers during a 2-year period, and Craig et al3 reported a 32% consultation rate during 1 year in a university-based family practice.
The major driving force for this variation is the restrictive privilege situation that exists in many Connecticut hospitals with regard to maternity care by family physicians. It is highly unlikely that differences in the risk status between patients in hospitals B and C could account for the wide difference in consultation rates. Privileging may well be a driving force. This privileging issue may further account for the fact that only 9.0% of family physicians in Connecticut deliver neonates.
In a national survey of family physicians, Wadland et al17 found that the scope of obstetric privileges varies by region. Nationally, approximately half of all family physicians who provide maternity care have required consultations for certain pregnancy-related conditions or procedures. The greatest freedom is found in the mountain plains states, where 60% of family physicians have unrestricted privileges, and the least freedom is in the East, where only 29% are unhindered by required consultations. The American Academy of Family Physicians must continue to support the efforts of its membership in obtaining and maintaining appropriate hospital privileges. Future research might examine perinatal outcomes in a randomized trial of mandatory vs voluntary obstetric consultation.
The neonatal outcomes in the study were good, but the data were very limited, since only Apgar scores were reported. Future investigations should include more information to determine the status of the newborn. Variables such as birth weight, cord pH values, need for resuscitation, need for prolonged hospital stay, and transfer to a neonatal intensive care facility might better define the health of the neonate.
Another weakness of the study was the inability to verify the information submitted on all of the obstetric data cards. The information is all self-reported by the delivering physician, with review of the labor and delivery log at only 3 of the institutions. Some physicians may have failed to complete cards following a delivery, leaving some potential gaps in the information. However, this technique of self-reporting has been used successfully by other research networks,18-19 and one should presume that the data reported by the participating family physicians are an accurate reflection of maternity care provided in Connecticut.
In summary, family physicians in Connecticut are important providers of maternity care to their patients. Despite limitations on their obstetric activities through restrictive hospital privileges, they continue to maintain their enthusiasm about participating in this important aspect of the family cycle. Some have called family physicians who deliver babies an "endangered species,"20 but to paraphrase Mark Twain, a famous Connecticut resident, reports of our death have been greatly exaggerated.21
AUTHOR INFORMATION
Accepted for publication January 28, 2000.
I thank John Cordes, MD, for his assistance in preparing the delivery data for his institution and Len Averill, MD, for facilitating access to the delivery log in his obstetric unit. I must also acknowledge all the members of the Connecticut Family Practice Obstetrics research network for their invaluable contributions to this project: Thomas Agresta, MD, Sarah Aronson, MD, Kevin Boucher, DO, David Boxwell, MD, Arthur Catsam, MD, Timothy Cooper, MD, John Cordes, MD, Joseph Creme, MD, Robert Cushman, MD, Margerite Davis, MD, Mary Guerrera, MD, Susan Hasti, MD, Nancy Hurlburt, MD, Michael Kazakoff, MD, Patricia Lampugnale, DO, J. Carey LaPorte, MD, Charlene Li, MD, Thomas McLarney, MD, Mary Helen Morrow, MD, Andrea Needleman, MD, James Nicholson, MD, Peter Rabinowitz, MD, Wendolyn Reymond, MD, Alberto Rodriguez, MD, Gregory Shields, MD, Keith Sinusas, MD, Gerald Sullivan, MD, Jane Walker, MD, Dael Waxman, MD, Olivia Wright, MD, Cathy Zack, MD, and Roy Zagieboylo, MD.
Reprints: Keith Sinusas, MD, Middlesex Hospital Family Practice Residency, 90 S Main St, Middletown, CT 06457.
From the Middlesex Hospital Family Practice Residency, Middletown, Conn.
REFERENCES
| |
1. Koning JH. The obstetrical experience of 20 years in one family practice. J Fam Pract. 1982;14:163-171.
PUBMED
2. Kriebel SH, Pitts JD. Obstetric outcomes in a rural family practice: an eight year experience. J Fam Pract. 1988;27:377-384.
ISI
| PUBMED
3. Craig AS, Berg AO, Kirkwood CR. Obstetric consultations during labor and delivery in a university-based family practice. J Fam Pract. 1985;20:481-485.
ISI
| PUBMED
4. Wanderer MJ, Suyehira JG. Obstetric care in a prepaid cooperative: a comparison between family practice residents, family physicians and obstetricians. J Fam Pract. 1980;11:601-606.
PUBMED
5. Hueston WJ, Applegate JA, Mansfield CJ, King DE, McCaflin RR. Practice variations between family physicians and obstetricians in the management of low-risk pregnancies. J Fam Pract. 1995;40:345-351.
PUBMED
6. Miser WF, Blount BW, LeClair BM, et al. The practice of obstetrics by army family physicians. J Am Board Fam Pract. 1996;9:174-181.
7. Smucker MD. Obstetrics in family practice in the state of Ohio. J Fam Pract. 1988;26:165-168.
PUBMED
8. Hueston WJ. Site-to-site variation in the factors affecting cesarean section rates. Arch Fam Med. 1995;4:346-351.
FREE FULL TEXT
9. Connecticut Hospital Association. Connecticut Health Information Management and Exchange Inc Database. Hartford: Connecticut Hospital Association; 1996.
10. Kozak LJ. Surgical and nonsurgical procedures associated with hospital delivery in the United States: 1980-1987. Birth. 1989;16:209-213.
PUBMED
11. Leveno KJ, Cunningham FG, Nelson S, et al. A prospective comparison of selective and universal electronic fetal monitoring in 34,995 pregnancies. N Engl J Med. 1986;315:615-619.
ABSTRACT
12. Kelso IM, Parsons RJ, Lawrence GF, et al. An assessment of continuous fetal heart rate monitoring in labor: a randomized trial. Am J Obstet Gynecol. 1978;131:526-532.
ISI
| PUBMED
13. Smith MA, Ruffin MT, Green LA. The rational management of labor. Am Fam Physician. 1993;47:1471-1481.
ISI
| PUBMED
14. Graham ID, Graham DF. Episiotomy counts: trends and prevalence in Canada, 1981/1982 to 1993/1994. Birth. 1997;24:141-147.
PUBMED
15. American Academy of Family Physicians. Family Practice in Health Care Organizations. Kansas City, Mo: American Academy of Family Physicians; 1996.
16. Hueston WJ. Obstetric referral in family practice. J Fam Pract. 1994;38:368-372.
PUBMED
17. Wadland WC, Havron AF, Garr D, Schneeweiss R, Smith M. National survey on hospital-based privileges in family practice obstetrics. Arch Fam Med. 1994;3:793-800.
FREE FULL TEXT
18. Klinkman MS, Stevens D, Gorenflo DW. Episodes of care for chest pain: a preliminary report from MIRNET. J Fam Pract. 1994;38:345-352.
ISI
| PUBMED
19. Miller RS, Iverson DC, Freid RA, Green LA, Nutting PA. Carpal tunnel syndrome in primary care: a report from ASPN. J Fam Pract. 1994;38:337-344.
PUBMED
20. Larimore WL, Reynolds JL. Family practice maternity care in America: ruminations on reproducing an endangered species—family physicians who deliver babies. J Am Board Fam Pract. 1994;7:478-488.
21. Beck EM, ed. Bartlett's Familiar Quotations. 15th ed. Boston, Mass: Little Brown & Co; 1980:625.
RELATED ARTICLE
The Archives of Family Medicine Continuing Medical Education Program
Arch Fam Med. 2000;9(5):463-465.
FULL TEXT
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
Electronic Data Collection Options for Practice-Based Research Networks
Pace and Staton
Ann Fam Med 2005;3:S21-S29.
ABSTRACT
| FULL TEXT
|
