International Journal for Quality in Health Care 16:327-332 (2004)
International Journal for Quality in Health Care vol. 16 no. 4 © International Society for Quality in Health Care and Oxford University Press 2004; all rights reserved
Effect of a clinical practice guideline on physician compliance
Obstetrics and Gynecology, Prince of Songkla University, Hatyai, Songkhla, Thailand
Objectives. To evaluate the effect of a clinical practice guideline for cesarean section due to dystocia on physician compliance, pregnancy outcome, and cesarean delivery rates, and to identify factors associated with physician non-compliance.
Design. A cross-sectional study.
Setting. A university hospital, Southern Thailand.
Study participants. All 719 medical records of women undergoing a cesarean section due to dystocia (failure to progress; cephalopelvic disproportion) before and after implementation of the guideline, from 1 January 1998 to 31 December 2000.
Intervention. A clinical practice guideline for cesarean section due to dystocia was implemented on 1 June 1999.
Main outcome measures. Physician compliance, pregnancy outcomes, and cesarean section rates. Multivariate logistic regression was used to identify factors associated with physician non-compliance. Independent variables consisted of maternal age, height, parity, type of service, and birthweight.
Results. Physician compliance with the guideline was 89.2%. Maternal complications were less in the period after implementation of the guideline. Fetal outcomes were not different between the two periods. The cesarean section rates due to dystocia decreased after implementation of the guideline, from 10.7% in 1999 to 8.6% in 2002. Private practice, maternal short stature, and birthweight 
3500 g were significant predictors of physician non-compliance.
Conclusions. Physician compliance was high. A clinical practice guideline can reduce the cesarean section rates due to dystocia without increasing adverse outcomes. Physician non-compliance was more common in women with well known risk for cephalopelvic disproportion, and private practice.
Keywords: audit, cesarean delivery, clinical practice guidelines, compliance
Address reprint requests to Chitkasaem Suwanrath-Kengpol, Obstetrics and Gynecology, Prince of Songkla University, Hatyai, Songkhla, Thailand. E-mail: schitkas{at}ratree.psu.ac.th or schitkas{at}hotmail.com
Accepted for publication April 12, 2004.
The overall cesarean section rate has increased progressively in many parts of the world, including Thailand, in particular over the past 10 years. This rising cesarean rate has become of increasing concern to both the obstetric profession and the public. There are many factors that contribute to variations in cesarean section rates, both medical and non-medical, such as practice culture, practice style, hospital environment, source of payment, patient preferences, and socioeconomic status [1–4]. Several strategies have been proposed to reduce cesarean rates, such as physician and public education about maternal and fetal benefits of vaginal delivery, practice guidelines for the management of labor, commitment of attending physicians to a lower cesarean delivery rate, auditing by peer review, intensive feedback of the outcomes, public dissemination of cesarean section rates, modification of physician reimbursement, modification of hospital reimbursement, and medical malpractice reform [1,5–7].
Our institution, Songklanagarind Hospital, Faculty of Medicine, Prince of Songkla University, the only university hospital in Southern Thailand, also has had a high cesarean section rate. From 1990 to 1998 it increased from 25% to 33%. In 1999, concerned with this trend, the Department of Obstetrics and Gynecology reached a consensus to reduce the cesarean delivery rates to improve the quality of maternal and fetal health care. Vaginal birth after prior cesarean (VBAC) is not done in our hospital due to limitations of medical personnel and resources, therefore we have mainly focused on reducing the primary cesarean section rate. Upon investigation, it was found that the major obstetric indication for primary cesarean section was dystocia (failure to progress; cephalopelvic disproportion), the diagnosis of which has progressively increased, with wide variations in diagnosis and management depending on individual obstetrician practices. In an attempt to reduce the increasing number of cesarean sections, a clinical practice guideline for cesarean sections due to dystocia was developed. It was derived and adapted from that of the Royal Thai College of Obstetricians and Gynecologists (RTCOG), with the agreement of all 18 obstetricians in Songklanagarind Hospital. Participation by attending obstetricians was voluntary and not linked to any sanction. It was decided to audit physician compliance and to evaluate the effect of the guideline after implementation to determine whether it was safe and effective in reducing the cesarean section rate.
The effect of the implementation of a guideline related to cesarean deliveries has been studied widely in many institutions. However, most studies have mainly focused on cesarean section rates and pregnancy outcomes [1,8–11]. There have been very few studies concerning physician compliance or actual practices, which are important factors in determining outcome [6,7]. No study that we are aware of has reported on factors associated with physician non-compliance. Therefore, we aimed at evaluating physician compliance with the clinical practice guideline, and to identify factors associated with physician non-compliance. In addition, we also evaluated the effect of the clinical practice guideline on pregnancy outcomes and rate of cesarean delivery. Our hypothesis was that physician compliance would be high, and cesarean sections due to dystocia could be decreased without adverse outcomes for either mothers or fetuses.
 |
Materials and methods |
|---|
 Top Materials and methods ResultsDiscussionReferences |
|---|
This study was undertaken in Songklanagarind Hospital, a university hospital located in Southern Thailand that serves as a regional tertiary center. There are
2500 deliveries per year. There are two types of service in the Department of Obstetrics and Gynecology: private and non-private. Private patients have their own obstetrician from ante- through postpartum. Non-private patients are cared for by resident physicians in obstetrics and gynecology, supervised by experienced obstetricians from the department.
The guideline was implemented into clinical practice on a voluntary basis on 1 June 1999. All resident physicians training in obstetrics and gynecology and all obstetricians committed to follow the guideline. Before making a decision to perform a cesarean section based on an indication of dystocia, the criteria for diagnosis of dystocia had to be fulfilled, which are: (i) cervical dilatation 3 cm; (ii) good uterine contraction for at least 2 hours (interval of 2.5–3 minutes, duration of 40–60 seconds, and intensity of at least 2+ by clinical assessment); and (iii) abnormal progression of labor, including one of the following: protraction disorder (<1.2 cm/hour in nulliparous and <1.5 cm/hour in multiparous), arrest disorder (arrest of dilatation for at least 2 hours or arrest of descent for at least 1 hour), or prolonged second stage of labor (2 hours in nulliparous and 1 hour in multiparous). If one of these conditions is not met, two obstetricians are required for the decision to proceed with the cesarean section. The Songklanagarind guideline was quite similar to that of the RTCOG, except for the criterion of cervical dilatation, which was changed from 4 cm with 80% effacement to only 3 cm, as we were concerned about the safety of the RTCOG guideline, and wanted to make it less stringent.
The required sample size was calculated based on the estimated prevalence of physician compliance with a clinical practice guideline. Setting a confidence interval of 95%, and an estimated prevalence of physician compliance from a pilot study of 85% with an acceptable error of 5%, indicated that at least 196 cases of cesarean section due to dystocia were required to estimate the prevalence of physician compliance reliably.
We reviewed all medical records of women with a singleton pregnancy who had undergone cesarean deliveries due to dystocia before and after implementation of the guideline, at Songklanagarind Hospital from 1 January 1998 to 31 December 2000.
Maternal characteristics were compared between the two periods using chi-square test and Student’s t-test. A P-value of <0.05 was considered significant. Data were derived from the database of the Statistical Unit of the Department of Obstetrics and Gynecology, Songklanagarind Hospital.
Physician compliance with the guideline, defined as practice that followed the protocol, was evaluated from medical records. We evaluated the effect of the guideline on pregnancy outcomes by using data in the earlier period, before implementation of the guideline for comparison.
The cesarean trend was analyzed from the cesarean section rates for the years 1990–1999 by a time series analysis using Holt’s exponential smoothing model. Time series analysis was used to identify the nature of the phenomenon represented by the sequence of observations, and to predict future values of the time series variable. Holt’s exponential smoothing model is a method of forecasting based on a simple statistical model of a time series. It is suitable for data with an explicit linear trend. Recent observations are given relatively more weight in forecasting than older observations. The Holt’s exponential smoothing model estimates a trend (T) component and a smoothed value (S) by the estimating equations: St = 
Xt + (1 – )(St–1 + Tt–1) and Tt = 
(St – St–1) + (1 – )Tt–1. and are smoothing constants with values between 0 and 1. Forecasts are calculated as Ft+m = St + Ttm (m = 1, 2, 3,...) and Ft+m is a predicted value at time t + m. We used the program to automatically set the appropriate and values. It was found that an value of 1.0 and a value of 0 produced the smallest sum of squared error in the fit model. Predicted cesarean rates after implementation of the guideline through the year 2002 were then created from the fit equation, including predicted rates for the years 2000, 2001, and 2002. The observed cesarean rates due to dystocia in the years 2000 to 2002 were then added in the data file. To demonstrate the effect of the guideline on cesarean delivery rates, a graph for the observed and predicted cesarean rates for the years 1990–2002 was constructed.
In the analysis for predictors of physician non-compliance, the potential predictor variables consisted of maternal age, parity, maternal height, type of service, and birthweight. They were all included in the multiple logistic regression model because one variable may have a confounding effect on the others. For fetal birthweight, we selected 3500 g as the threshold of abnormal weight instead of 4000 g because Thai babies are normally small; the average fetal birthweight in our hospital is 3000 g. The average birthweight of cesarean babies due to dystocia is approximately 3300–3400 g. SPSS for Windows version 10.0 was used for data analysis. A P-value of <0.05 was considered significant.
|
Results |
|---|
|
TopMaterials and methodsResultsDiscussionReferences |
|---|
From 1 January 1998 to 31 December 2000, there were 3126 and 3768 deliveries before and after implementation of the guideline, respectively. Demographic characteristics were not notably different between the two periods (Table 1). A total of 719 women had cesarean delivery due to dystocia. There were 320 and 399 cases in the periods before and after implementation of the guideline, respectively.
|
Physician compliance with the clinical practice guideline evaluated from all 399 medical records was 89.2%, with 83.1% and 98.7% in private and non-private practices, respectively. Physician compliance was significantly different between the two groups (P < 0.001). No obstetricians followed the criterion that requires two obstetricians to make the decision. The criteria of cervical dilatation 3 cm, good uterine contraction, and abnormal progression of labor were not followed in 5.8%, 5.8%, and 4.8% of cases, respectively. The majority of cases had arrest disorder (87.7%). In the non-compliance group, 41 out of 43 cases (95.3%) were private patients. In addition, the median cervical dilatation at the time of diagnosis of dystocia in this group was only 2 cm (range 1–10 cm).
Pregnancy outcomes were compared between the two periods. Postpartum complications including uterine atony, metritis/endometritis, postpartum hemorrhage, puerperal morbidity, and wound infection were less in the period after implementation of the guideline. Fetal outcomes, including birthweight, Apgar scores, meconium-stained amniotic fluid, and admission to a neonatal intensive care unit, were not different (Table 2). No cases had an Apgar score at 5 minutes of <7. No cases had cesarean hysterectomy. No fetal or maternal deaths or any serious complications were reported. Comparison of the pregnancy outcomes of all deliveries between the two periods using the database of the department showed that there were no significant differences in terms of postpartum complications and perinatal mortality rates. Postpartum complications were 3.5% and 3.3%, and the perinatal mortality rates were 7.8/1000 and 8.4/1000 in the periods before and after implementation of the clinical practice guideline, respectively.
|
The time series analysis showed that the trend of cesarean deliveries due to dystocia was rising, with a predicted cesarean rate of 12.7% by the year 2002. However, after implementation of the guideline, the observed cesarean rate fell from 10.7% in 1999 to 8.6% in 2002, deviating from the prediction (Figure 1). During the same period, departmental data showed that the primary cesarean section rate fell from 22.3% in 1999 to 21.2% in 2002, while the cesarean section rate due to fetal distress did not increase significantly (3.3% to 3.8%).
|
It was found that private practice, nulliparity, and birthweight 3500 g were significantly associated with physician non-compliance upon univariate analysis (Table 3). In the multivariate logistic regression analysis, the factors significantly associated with physician non-compliance were private practice, maternal short stature, and high birthweight. Private practice was a very strong predictor of physician non-compliance compared with non-private practice. Nulliparity and maternal age were not significant predictors in the multivariate logistic regression model (Table 4).
|
|
|
Discussion |
|---|
|
TopMaterials and methodsResultsDiscussionReferences |
|---|
In this study, it was found that physician compliance with the clinical practice guideline for cesarean delivery due to dystocia was quite high, although it did not reach 100% overall, perhaps because the policy was introduced on a voluntary basis. Compliance in non-private practice was significantly higher than in private practice.
Fetal outcomes did not differ between the two periods. There were no maternal or fetal deaths in either study group. In addition, postpartum complications were less in the period after implementation of the guideline. However, this might occur by chance because of the small sample size. Departmental data showed that pregnancy outcomes of all deliveries between the two periods were not significantly different, possibly indicating that the guideline reflects safe and good practice. In this study, we only examined deviation from the guideline by overuse, not instances of underuse.
The cesarean section rate due to dystocia was reduced after implementation of the clinical practice guideline, indicating that the guideline was effective in lowering the rate of cesarean deliveries. In addition, the primary cesarean section rate was slightly reduced and the cesarean section rate due to fetal distress did not significantly increase during the same period. Therefore, the reduction of the cesarean rate for dystocia was not a shift in the primary cesarean indication. Other institutions have had considerable success in reducing cesarean rates with guideline implementation [8,9,12], along with other strategies including service improvements, peer review, and individual feedback [1,7].
Multivariate logistic regression analysis indicated that private practice had 16 times the rate of physician non-compliance compared with non-private practice, probably because management in private practice is more based on individual judgment, patient preference, and financial factors. Maternal short stature and high fetal birthweight were also significant factors associated with physician non-compliance; both are well-known risk factors of cephalopelvic disproportion and might affect the obstetrician’s decision to perform a cesarean section earlier, in the latent phase or before diagnostic criteria have been confirmed. However, no obstetricians followed the criterion that requires two obstetricians to make such a decision, indicating that the policy is not strictly enforced. We recommend that these women be managed cautiously. If the guideline is followed too strictly, it could be harmful to both mother and fetus. The criterion that requires two obstetricians for decision approval should be considered in such cases.
Our hospital was successful in reducing the rate of cesarean deliveries due to dystocia by implementation of the clinical practice guideline with no adverse outcomes. We believe that initiative within the department, obstetricians’ participation in making the policy, and their self-commitment to follow the guideline were important factors in this success. However, the cesarean section rate due to this indication in our institution is still high. This may be because our guideline is not stringent, or obstetrician compliance, particularly in private practice, is still low, or because the evaluation period was not long enough to reach the desired rate. The appropriate rate should be 4.5% of total deliveries, estimated from the recommended cesarean section rate for all indications by the World Health Organization of 15%, and this indication accounts for 30% of total cesarean deliveries in our hospital. The available literature indicates that effective strategies include: advocacy by leadership and building consensus; individual participation in policy making (compared with mandated programs); consistent individual feedback; readiness to change; and non-blinded, intradepartmental distribution of outcomes [13–15]. Widely used continuing medical education delivery methods such as conferences apparently have little direct impact on improving professional practice [13].
We have proposed more strategies in an attempt to decrease further cesarean deliveries due to dystocia to an acceptable rate, including: adaptation of the guideline to be more stringent by changing the criterion of cervical dilatation to 4 cm with 80% effacement, as specified by RTCOG; use of a check list of criteria in the guideline for every cesarean delivery due to dystocia with audit by peer review; analysis of obstetricians’ individual cesarean rates; and intensive feedback of the outcomes by comparing with others, especially the outliers. All obstetricians have been informed that their practice will be monitored. We hope that not only the rate of cesarean section due to dystocia decreases, but also the overall primary cesarean delivery rate. Since our institution is a medical school, performing cesarean sections with inappropriate indications, such as patient’s desires or elderly primigravida, is not permitted. We will again evaluate cesarean section rates in the future to see if they have significantly decreased to an acceptable rate through the new strategies.
|
References |
|---|
|
TopMaterials and methodsResultsDiscussionReferences |
|---|
- Poma PA. Effect of departmental policies on cesarean delivery rates: a community hospital experience. Obstet Gynecol 1998; 91: 1013–1018.[Medline]
- Berkowitz GS, Fiarman GS, Mojica MA, Bauman J, de Regt RH. Effect of physician characteristics on the cesarean birth rate. Am J Obstet Gynecol 1989; 161: 146–149.[Web of Science][Medline]
- Guillemette J, Fraser WD. Differences between obstetricians in cesarean section rates and management of labour. Br J Obstet Gynaecol 1992; 99: 105–108.[Web of Science][Medline]
- Olatunbosun A, Ravichander A, Turnell RW, Edouard L. The influence of patient preferences and physician practices on cesarean delivery. Clin Exp Obstet Gynecol 2002; 29: 19–21.[Medline]
- Stafford RS. Alternative strategies for controlling rising cesarean section rates. JAMA 1990; 263: 683–687.
[Abstract/Free Full Text] - Lomas J, Enkin M, Anderson GM, Hannah WJ, Vayda E, Singer J. Opinion leaders vs audit and feedback to implement practice guidelines. Delivery after previous cesarean section. JAMA 1991; 265: 2202–2207.
[Abstract/Free Full Text] - Wareham V, Bain C, Cruickshank D. Cesarean section audit by peer review. Eur J Obstet Gynecol Reprod Biol 1993; 48: 9–14.[Medline]
- Iglesias S, Burn R, Saunders LD. Reducing the cesarean section rate in a rural community hospital. Can Med Assoc J 1991; 145: 1459–1464.[Abstract]
- de Muylder X, Thiery M. The cesarean delivery rate can be safely reduced in a developing country. Obstet Gynecol 1990; 75: 360–364.[Medline]
- Myers SA, Gleicher N. A successful program to lower cesarean section rates. N Engl J Med 1988; 319: 1511–1516.[Abstract]
- Socol ML, Garcia PM, Peaceman AM, Dooley SL. Reducing cesarean births at a primarily private university hospital. Am J Obstet Gynecol 1993; 168: 1748–1754.[Medline]
- Zanetta G, Tampieri A, Currado I et al. Changes in cesarean delivery in an Italian university hospital, 1982–1996: a comparison with the national trend. Birth 1999; 26: 144–148.[CrossRef][Medline]
- Davis DA, Thomson MA, Oxman AD, Haynes RB. Changing physician performance. A systematic review of the effects of continuing medical education strategies. JAMA 1995; 274: 700–705.
[Abstract/Free Full Text] - Main EK. Reducing cesarean birth rates with data-driven quality improvement activities. Pediatrics 1999; 103: 374–383.
- Robson MS, Scudamore IW, Walsh SM. Using the medical audit cycle to reduce cesarean section rates. Am J Obstet Gynecol 1996; 174: 199–205.[Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||