Do provider birth attitudes influence cesarean delivery rate: a cross-sectional study

Do provider birth attitudes influence cesarean delivery rate: a cross-sectional study Background: When used judiciously, cesarean sections can save lives; but in the United States, prior research indicates that cesarean birth rates have risen beyond the threshold to help women and infants and become a contributor to increased maternal mortality and rising healthcare costs. Healthy People 2020 has set the goal for nulliparous, term, singleton, vertex (NTSV) cesarean birth rate at no more than 23.9% of births. Currently, cesarean rates vary from 6% to 69% in US hospitals, unexplained by clinical or demographic factors. This wide variation in cesarean use is also seen among individual providers of intrapartum care. Previous research of birth attitudes found providers of intrapartum care hold widely differing views, which may be a key underlying factor influencing practice variation; however, further study is needed to determine if differences in attitudes are associated with differences in clinical outcomes. The purpose of this study was to estimate the association between individual provider attitudes towards birth and their low-risk primary cesarean rate. Methods: Four hundred providers were drawn from a stratified random sample of all California providers of intrapartum care in 2013 and surveyed for their attitudes towards various aspects of labor and birth. Providers’ NTSV cesarean birth rates were obtained for 2013 and 2014. Covariates included gender, years of experience, practice location, and primary hospital’s NTSV cesarean rate. We used adjusted multivariate Poisson regression to compare cesarean rates and linear regression to compare attitude scores of providers meeting versus not meeting the Healthy People 2020 (HP2020) goal. Results: Two hundred nine total participants (obstetricians, family physicians, and midwives) completed surveys, of which 109 perform cesareans. Providers’ NTSV cesarean rate was significantly associated with their composite attitudes score [IRR for each one-point increase 1.21 (95% CI 1.002–1.45)]. Physicians meeting the HP2020 goal held attitudes which were significantly more favorable towards vaginal birth: mean 2.70 (95% CI 2.58–2.83) versus 2.91 (95% CI 2.82–3.00), p < 0.01. Conclusions: Provider attitudinal differences are associated with NTSV cesarean rates. Those meeting the HP2020 goal hold attitudes more favorable towards vaginal birth. These findings may present a modifiable target for quality improvement initiatives to decrease low risk primary cesareans. Keywords: Cesarean section, Provider attitudes, Quality improvement, Primary cesarean, Culture * Correspondence: EWhiteVanGompel@northshore.org Department of Family Medicine, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, Evanston, IL 60201, USA Department of Obstetrics and Gynecology, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, Evanston, IL 60201, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 2 of 9 Background service of the California Maternal Quality Care Collab- Cesarean sections, when used judiciously, save lives; orative. Baseline demographics and survey scores were however, research indicates that the use of cesareans in analyzed using SAS Version 9.4. Poisson regressions were the United States has risen well above the level of neces- performed using Stata 12.1. This study was approved sity and has become a contributor to maternal morbidity under expedited review by University of California Davis and mortality [1–5]. Women who are nulliparous, at full Internal Review Board. Participants received the consent term, with a singleton pregnancy in vertex presentation document, but signed consent was waived by the IRB. (NTSV) have been established as a standard population and used as a target group for reducing the cesarean Setting birth rate [6–8]. Healthy People 2020 has set the goal California has approximately 3500 providers that are for NTSV cesarean rate at no more than 23.9% of births listed as delivering clinicians in a given year; these [6]. Currently, cesarean rates vary from 6% to 69% in US include obstetricians (OB), Maternal-Fetal Medicine hospitals [9]. Substantial variation persists after adjust- specialists, family medicine physicians, Certified Nurse ment for hospital demographics, referral categories, or Midwives (CNM), and California Licensed Midwives teaching status, and after adjustment for patient clinical (LM). Practice settings are diverse, and include rural, or sociodemographic factors [9, 10]. urban, and frontier geographic locations, teaching hospi- Wide variation is also seen between providers, even tals, community hospitals, and tertiary care centers. Pro- those practicing within the same hospital and utilizing a vider mix at each hospital may include only obstetricians laborist model [11]. Over 90% of the variation in the or family physicians, or may include Maternal-Fetal NTSV cesarean rate is due to two indications: “fetal in- Medicine specialists or midwives. The CMDC combines tolerance of labor” and “failure to progress”; indications existing datasets, including monthly discharge and requiring subjective decision-making by the intrapartum clinical data, birth certificate data, and semi-annual patient provider [12]. The American Congress of Obstetricians dischargedatafromthe California Office of Statewide and Gynecologists’ Committee Opinion on limiting Health Planning and Development, to create hospital and intervention during labor highlights the labor manage- physician-level quality improvement metrics. We obtained ment techniques that, despite prevailing evidence, vary a stratified random sample from 2013 - the latest available significantly among providers [13]. Evidence is filtered year of complete data. At the time of final data collection, through the lens of a provider’s experiences and 2014 complete data became available, thus we included attitudes [14]; yet how these attitudes affect clinical out- provider and hospital metrics from both 2013 and 2014 comes has not been evaluated. combined. Our primary sample size consideration was Previous research found differences in birth attitudes based on the six domain scores’ ability to account for vari- between providers of different disciplines (obstetrics, ation in NTSV cesarean delivery rate. Using pilot data, we family medicine, midwifery) [15, 16]; yet did not exam- determined that a sample size of 116 patients would ine clinical outcomes. The objective of this study was to provide 80% power to detect an 11.1% partial R-square. To examine the association of individual provider NTSV account for intermittent missing variables, we targeted a cesarean rates with their attitudes towards birth. In con- sample of 130 survey participants. trast to earlier studies examining physician demographic Inclusion criteria for providers were: having been listed factors [17–19], this study sought to evaluate a poten- on a birth certificate as the delivering clinician in the tially modifiable personal attribute that may contribute year 2013, belonging to one of the key study disciplines to cesarean overuse. (maternal-fetal medicine, OB, family medicine, CNM, LM), and having performed at least 10 deliveries per Aim year. Providers with a license address outside the state of Our primary aim was to assess the association between California and those without an identifiable discipline providers’ birth attitudes and their NTSV cesarean rates. based either on license number prefix or NPI taxonomy code were excluded from sampling. Methods Based on prior research pointing to important influ- Design ences on provider practice [17–19], stratification was A stratified random sample of providers listed as deliver- performed based on three variables: provider discipline ing clinicians on California birth certificates in 2013 (maternal-fetal medicine, OB, family medicine, CNM, were surveyed with a previously validated survey instru- LM), geographic location as defined by the federally des- ment of provider birth attitudes [16]. We linked pro- ignated Medical Service Study Areas (rural = population viders’ scores on this survey with their individual NTSV density < 250 persons/square mile; frontier = population cesarean rates for the years of 2013 and 2014, as calcu- density < 11 persons/square mile; urban = anything not lated by the California Maternal Data Center (CMDC), a rural or frontier) [20], and years in practice (< 5 years, White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 3 of 9 5–15 years, 16–25 years, > 25 years). The stratified ran- a provider that qualify as primary low-risk cesareans. This dom sample was drawn in two rounds of 200, for a total criteria includes all nulliparous, term (> 37 weeks gestation), of 400 sampled providers. To ensure adequate power for singleton gestations in vertex presentation at delivery. These analysis, oversampling was performed on provider are termed NTSV births for nulliparous, term, singleton, groups other than obstetricians (who perform the vast and vertex, and remain the primary target for quality im- majority of all births in California). Researchers were provement initiatives to decrease cesarean section overuse blinded to individual providers’ cesarean rates during in the United States [6–8]. We obtained both the total num- sample selection. ber of NTSV births providers attended in 2013–2014, and the total number of those births that were delivered via Attitudes survey cesarean from the CMDC database. The birth attitudes survey, previously validated by Klein Not all providers included in the survey sample had privi- and colleagues, included 9 different domains assessing pro- leges to perform cesareans. In order to provide a complete vider attitudes towards different aspects of labor and birth picture of the spectrum of attitudes held by all providers [16]. Six of these domains, comprising 31 Likert-style items, practicing as independent clinicians in California, we chose were chosen as thosemostlikelytohaveaneffect on the to include all those listed as delivering clinicians on at least targeted outcome – low-risk primary cesareans. They in- 10 deliveries per study year. As it would not be appropriate cluded (renumbered for this study): Domain 1: attitudes re- to assess provider-level cesarean rates in providers who garding use of electronic fetal monitoring (Cronbach alpha cannot perform cesareans, we excluded these providers [α] = 0.704), Domain 2: factors that increase cesarean rates from theNTSVpredictivemodel.BothCNMsand LMsdo (α = 0.810), Domain 3: fears of birth mode by respondents not perform cesareans under any circumstances. A majority or their partners/spouses (α=0.929),Domain4:factorsthat of family medicine physicians who do prenatal and intra- decrease cesarean rates (α = 0.819), Domain 5: maternal partum care do not perform cesareans; however, there are choice and mothers’ roles in birth (α = 0.646), and Domain family medicine physicians who have done additional train- 6: safety by mode or place of birth (α = 0.748). The compos- ing to qualify to perform cesareans. There is no central ite scale combined the individual domain scores. Coding of database that tracks these privileges, thus this was ascer- domains 4 and 5 was reversed for directional consistency to tained by anonymously calling individual family medicine create a total mean score. Lower scores on the composite physicians’ offices. scale indicate attitudes more favorable toward vaginal birth, while higher scores indicate attitudes that favor cesarean Covariates birth. Providers were also asked to give their discipline and Provider demographic data, including discipline (mater- years of experience. Each provider was assigned a random nal-fetal medicine, Average Risk OB, family medicine, identifier, so that survey responses were not associated with CNM, LM), gender, years since graduation from medical provider name. school (< 5 years, 5–15 years, 16–25 years, > 25 years), We used best practices for achieving maximal survey practice geographic location (rural/urban), and hospital- response as detailed by Dillman et al. [21] Survey data level demographics including hospital-level NTSV rate, collection began in October 2015 and ended in April neonatal intensive care unit level, and percent Medicaid 2016. We sent the attitudes survey via postal mail and were obtained from the CMDC database. Some providers included a web address for optional completion online. practiced at multiple hospitals; however, the hospital The initial mailing included the 4-page survey, a cover where they had their greatest number of NTSV births was letter, informed consent document, a self-addressed assigned as their primary hospital of delivery. stamped envelope, and an incentive of a $10 Starbucks gift card. Providers could opt to provide their email Statistical methods addresses to be entered into a drawing at the conclusion Due to the large range of providers’ total NTSV birth vol- of the study for an iPad of approximately $400 value at ume, we used multivariate Poisson regression to maximize the conclusion of the study. One to two weeks after the precision of our estimation model, using counts of NTSV initial mailing, we sent a postcard reminder to complete cesarean deliveries per provider as the dependent variable the survey. Between one and 2 months after the initial and total deliveries per provider as an “exposure” variable mailing, all non-responders were sent a second complete whose log-transformed value is included as an offset term, packet including cover letter, consent, survey document, to account for between-provider variation. We used and self-addressed stamped envelope. robust standard error estimators to protect against model misspecification. Simple regression with predicted margins Provider-level cesarean rate was used to compare attitudinal scores between discipline The CMDC database uses the standard NQF-endorsed algo- groups. Attitude domain and composite means were cal- rithm [4] to calculate the total number of births attended by culated from all non-missing items. For the dichotomous White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 4 of 9 Healthy People 2020 provider comparison, we compared had higher annual birth volumes than non-respondents’ providers’ attitudes based on whether or not their NTSV hospitals (mean = 5,203.9 versus 2,666.8, p < 0.001). Hos- cesarean rates met the Healthy People 2020 (HP2020) goal pital NTSV rates were slightly lower for respondents com- of less than 23.9% [6]. We used multiple linear regression, pared with non-respondents (25.3% versus 26.8%, p < 0.01). adjusting for provider gender, practice geography, and (Table 1). experience level to compare predicted means for each Attitudes varied significantly according to provider group. In order to adjust for the provider’shospital discipline (Fig. 1). Each domain displayed a spectrum of cultural environment, we adjusted each regression model provider attitudes towards birth, with midwives on one for the primary hospital’s NTSV rate exclusive of pro- end of the spectrum and OBs on the other (Fig. 2). Fam- vider’scontributionto this rate. ily medicine physicians either fell between the two or held attitudes more similar to midwives. Maternal-fetal Results medicine physicians held attitudes most consistent with We received 209 completed surveys, a total response rate average risk OBs, but tended to express attitudes that of 52.3%, including 22 maternal-fetal medicine, 101 OB, 53 fell between average risk OBs and family medicine physi- family medicine, 16 CNM, and 17 LM responding. There cians. The OB group had widest variation in composite was a higher response rate for midwives, 97% of who were attitude scores (range = 1.37 to 4.33). female, but no difference in disciplinary distribution or Of the 53 responding family medicine physicians, only gender of physician respondents versus non-respondents. 8 had privileges to perform cesareans, though 22 had Experience level, practice geographic location, and primary “first assist” privileges. Family medicine physicians with hospital nursery level did not vary between respondents cesarean privileges (family medicine-CS) were identical versus non-respondents. Respondents’ primary hospitals to family medicine physicians without privileges on the Table 1 Demographic and Practice Characteristics of Responders and Non-responders Responders (n) Percent Non-Responders (n) Percent p-value N 209 191 Discipline 0.35 MFM 22 10.5 19 9.9 OB 101 48.3 120 62.8 Family Medicine 53 25.4 46 24.1 CNM 16 7.7 3 2.1 LM 17 8.1 3 1.6 Gender 0.29 Female 132 63.2 102 53.4 Male 77 36.8 89 46.6 Experience 0.49 < 5 years 35 18.3 28 13.9 5–15 years 58 30.4 70 34.7 16–25 years 46 24.1 43 21.3 > 25 years 52 27.2 61 30.2 Hospital NICU Level 0.64 Basic Nursery 50 23.9 57 29.8 Community Nursery 82 39.2 73 38.2 Intermediate Nursery 36 17.2 33 17.3 Regional Nursery 20 9.6 25 13.1 Continuous Variables Mean SD Mean SD p-value Hospital Birth Volume 5203.9 3835.9 2666.8 2012.8 < 0.0001 Hospital NTSV CS Rate 25.30 4.80 26.80 6.10 < 0.01 Abbreviations: MFM maternal fetal medicine, OB obstetricians, Family Medicine family medicine physicians, CNM certified nurse midwives, LM licensed midwives, NICU neonatal intensive care unit, NTSV CS nulliparous, term, singleton, vertex cesarean section P-value includes physicians only. Midwives were all female except for one participant and had exceptionally high response rate Calculated for provider’s primary delivering hospital site White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 5 of 9 Fig. 1 Mean Provider Attitudes Scores on the Composite Scale by Provider Discipline. Figure 1 is a boxplot of providers’ mean attitude scores on the composite attitudes scale categorized by provider membership in a training discipline. Abbreviations: MFM: maternal fetal medicine; OB: obstetricians; family medicine: family medicine physicians; CNM: certified nurse midwives; LM: licensed midwives composite scale (mean 2.31 with SD 0.16, and 2.31 with providers meeting this goal and 65% of providers SD 0.35, respectively, p = 0.98). For the most part, family whose rate exceeded the goal. Providers meeting the medicine-CS physicians expressed attitudes on the individ- HP2020 goal held attitudes more favorable toward ual domains that were slightly less favorable towards vagi- vaginal birth compared with those over the HP2020 nal delivery than family medicine physicians; however, goal [adjusted mean 2.70 (95% CI 2.58–2.83) versus family medicine-CS tended to endorse less fear of vaginal 2.91 (95% CI 2.82–3.00), p < 0.01]. delivery than their family medicine counterparts, though this was not statistically significant (mean 1.19 ± 0.30 versus Discussion 1.40 ± 0.74, respectively, p = 0.43). The database attributed We found a significant association between providers’ cesarean deliveries to 14 family medicine physicians with- attitudes and beliefs about birth to their own NTSV out any cesarean privileges and 10 family medicine physi- cesarean rate. This study was consistent with the find- cians with first assist privileges. ings of the original Canadian survey validation study The analyses of association of provider attitudes with [16], finding that California providers’ attitudes towards NTSV cesarean rates included only providers with cesarean birth were primarily divided along disciplinary lines. privileges and at least 20 NTSV deliveries within the 2 year One of the most surprising findings in the disciplinary study period: 11 maternal-fetal medicine, 91 OB, and 7 analysis was the wide range of obstetricians’ attitudes, family medicine physicians. The composite attitudes scale spanning views more pro-cesarean than maternal-fetal scores for these providers ranged from 1.73 to 4.33 (mean 2. medicine providers all the way to those consistent with 83, standard deviation 0.48). Adjusted incidence rate ratios midwives. This large variation within a single discipline for domain and composite scale from the Poisson multiple suggests that there may be key acculturation differences regression are shown in Table 2; this model adjusted for in obstetric training and practice environment, where a provider gender, practice geography, experience level, and provider’s attitudes and beliefs are influenced. primary hospital’s NTSV rate (exclusive of provider’scontri- We took prior work an important step further by con- bution to this rate). The composite attitudes scale had an necting provider attitudes to their own measured clinical incident rate ratio of 1.21 (95% CI 1.002–1.45, p = 0.048), outcomes. For providers with cesarean privileges, those indicating that, for every 1 point increase in a provider’s meeting the Healthy People 2020 goal held attitudes that score on the composite scale, their NTSV cesarean rate were more favorable toward vaginal birth than those not decreased relatively by 21%. For individual domains, meeting this goal. In our regression analysis, as attitudes attitudes towards the perceived benefits of cesarean and fear became more favorable toward cesarean and less favor- of vaginal birth approached significance [respectively, 1.07 able toward vaginal birth, a provider’s NTSV cesarean (95% CI 0.99–1.17) and 1.08 (95% CI 0.99–1.17)]. (Table 2). rate increased proportionately. Additionally, we used a When dichotomizing providers by the Healthy People novel adjustor to account for local hospital culture – the 2020 NTSV cesarean goal cut-off of 23.9%, our sample primary hospital’s NTSV cesarean rate less the provider’s reflected the overall California distribution, with 35% of contribution to that rate. This adjustment highlights the White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 6 of 9 Fig. 2 a-f Boxplots of Provider Attitudes Scores on the Individual Domain Scales by Provider Discipline. Figure 2 a-f includes individual boxplots of providers’ mean attitude scores for each individual attitudinal domain categorized by provider membership in a training discipline. Abbreviations: MFM: maternal fetal medicine; OB: obstetricians; family medicine: family medicine physicians; CNM: certified nurse midwives; LM: licensed midwives White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 7 of 9 a b Table 2 Adjusted Associations between Attitude Scores and Provider NTSV CS Rate Domain Incidence Rate Ratio 95% CI p-value Fetal Monitoring 1.01 0.94–1.10 0.71 Benefits of CS 1.07 0.99–1.17 0.10 Fear of Vaginal Birth 1.08 0.99–1.17 0.09 Measures to Decrease CS 0.91 0.77–1.08 0.28 Maternal Role in Birth 0.93 0.85–1.02 0.14 Safety by Place or Mode of Birth 1.12 0.97–1.30 0.14 Composite Scale 1.21 1.002–1.45 0.048 Abbreviations: NTSV CS nulliparous, term, singleton, vertex cesarean section, CS cesarean section, CI confidence interval Results are from individual Poisson regression models of the NTSV cesarean outcome, with one model for each row, with that variable the focal predictor and with additional covariates used to adjust for gender, experience level, geographic location of practice, primary hospital’s NTSV cesarean rate calculated without the individual provider’s contribution Only includes providers with confirmed privileges to perform cesarean sections who had at least 20 NTSV births over the two-year study period of 2013–2014 Higher scores indicate attitudes more favorable toward cesarean section except for the two reverse-coded scales (as below) Higher scores indicate attitudes more favorable toward vaginal birth on these scales impact of individual provider attitudes on top of that of program significantly decreased their primary cesarean local hospital norms and practices. These findings sug- rate by providing senior obstetric supervision of resi- gest that a provider’s underlying attitudes, values, and dents on labor and delivery, highlighting the impact of beliefs play an important role in intrapartum decisions preceptor experience level on trainees [26]. The impact that ultimately affect birth outcomes. of integrating midwives into traditional obstetric training Our results are consistent with, and provide a possible has been posited but not yet rigorously tested against underlying mechanism for, other studies that have found clinical outcomes [27]. Finally, ongoing training and provider differences such as demographics, litigation support after experiencing a traumatic delivery event history, and practice environment are associated with may mitigate some of the fear attitudes associated with provider self-report personal thresholds to perform increased cesarean rate, which appear to impact entire cesareans [22, 23]. By independently measuring attitudes hospital units and not just the providers involved. This while protecting the confidentiality of providers and was described recently in a study of unplanned hospital using administrative data to associate attitudes with an cesarean rates, which increased and stayed elevated for unbiased measure of the individual provider-level 4 weeks after any catastrophic neonatal outcome within cesarean rate, our study minimized social desirability that hospital [28]. bias, which can confound self-report data. Of note, the actual magnitude of the point estimate of Limitations the association between composite attitudinal score and This cross-sectional study cannot draw conclusions NTSV cesarean rate is quite large. For every 1 point regarding causality or time course between independent increase in agreement with attitudes favoring cesarean, and dependent variables. Additionally, our a priori sam- NTSV cesarean rates increased by a relative 21%. For a ple size justification assumed a linear regression analysis, provider with a baseline rate of 25%, this would translate but we found that a Poisson regression analysis was bet- into an absolute change of 5.25%. In comparison, the ter suited to the outcome distribution. Thus, in order to QUARISMA trial found changes to the absolute assess the adequacy of the realized sample size for the cesarean rate of 0.7–2.3% resulting from an audit and reported effect sizes, one should consider the range of feedback mechanism in conjunction with hospital-based values included within the 95% CI [29]. Where the null best practices implementation [24]. Targeting attitudinal value was included along with values that would be change, which is a “bottom-up” method, in conjunction clinically meaningful, one could conclude that the effect with quality improvement, or “top-down”, efforts may is ambiguous and would require a more precise estimate enhance the impact of interventions. Research is needed in future work. For example, we would assert that a 15% to identify effective interventions to enhance evidence- relative increase in the NTSV cesarean rate would be based attitudes towards vaginal birth. particularly clinically meaningful. When examining the Training may offer a promising target for influencing 95% CI for the effect size of the individual attitude items, attitudes that favor vaginal birth. One study found that we find that all of them included the null value (i.e. were providers who trained in hospitals with lower obstetric not statistically significant) but three of them extended complication rates continued to have lower complication beyond an IRR of 1.15, suggesting that the results for rates once in practice [25]. Most recently, one hospital those items are ambiguous and could warrant a study with White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 8 of 9 a larger sample size. For the composite attitude meas- Medicine Jump Start Fund. Neither funding body had any role in the design, collection, analysis, interpretation of data, or in writing the manuscript. ure, we found statistically significant and clinically significant effects. Availability of data and materials Finally, the administrative data used to assess provider- The data that support the findings of this study are available from the California Maternal Data Center, but restrictions apply to the availability of these data, level cesarean rate was restricted to the provider listed on which were used under license for the current study, and so are not publicly the birth certificate and subject to errors of attribution. For available. Data are however available from the authors upon reasonable request example, the provider listed as “Delivering Clinician” on the and with permission of the California Maternal Data Center. Birth Certificate is usually but not always the person that Authors’ contributions managed the majority of a patient’slabor or made the EWV and JM designed the study with statistical consultation from DT and decision to go to cesarean. Unfortunately, this method content area expertise from EM. EM created the provider data file for sampling and supplied outcomes data for participants. EWV and DT analyzed discounts information about providers who do not carry and interpreted the attitudinal and outcomes data. EWV wrote the majority cesarean privileges but may play a major role in intrapar- of the manuscript with major contributions from all authors. All authors read tum care, and may be key decision-makers along the route and approved the final manuscript. that ends in either cesarean or vaginal delivery. For Ethics approval and consent to participate example, a provider may decide to admit a patient prior to This study was approved under expedited review by the University of the onset of active labor or use continuous electronic fetal California Davis Internal Review Board. Participants received the consent document, but signed consent was waived by the IRB. monitoring despite a patient’s low-risk status, both of which increase the likelihood of that patient requiring a cesarean, Competing interests yet the cesarean birth would be attributed to the clinician The authors declare that they have no competing interests. who performed the surgery itself and not the provider man- aging the labor. Further studies are needed to assess the Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in interplay of personnel on labor and delivery wards, how key published maps and institutional affiliations. decisions are made and birth outcomes attributed. Author details Department of Family Medicine, The University of Chicago, Pritzker School Conclusions of Medicine, NorthShore University HealthSystem Research Institute, 1001 This is the first study we are aware of linking provider attitu- 2 University Place, Evanston, IL 60201, USA. Department of Obstetrics and dinal differences to differences in measured birth outcomes. Gynecology, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, We found that the more providers’ birth attitudes favored Evanston, IL 60201, USA. California Maternal Quality Care Collaborative, cesarean, the higher their NTSV cesarean rate. In contrast Stanford University, Stanford Medical School Office Building, 1265 Welch to earlier studies that have focused on physician demograph- Road, MS 5415, Stanford, CA 94305, USA. Center for Healthcare Policy and Research and Department of Pediatrics, University of California Davis School ics or litigation history, which arerarelymodifiablewithout of Medicine, 2103 Stockton Blvd, Sacramento, CA 95817, USA. Center for major policy changes, we sought to evaluate the influence of Healthcare Policy and Research and Department of Family and Community a modifiable attribute – aprovider’s personal attitudes and Medicine, University of California Davis School of Medicine, 2103 Stockton Blvd, Sacramento, CA 95817, USA. beliefs. These findings suggest further scrutiny is needed of how future providers are acculturated during training and Received: 28 July 2017 Accepted: 19 April 2018 while in practice, and how practice groups, hospital units, and inter-professional interactions may modify these References attitudes to ultimately improve quality of care and health 1. Caughey AB, Cahill AG, Guise JM, Rouse DJ. Safe prevention of the primary outcomes. cesarean delivery. Am J Obstet Gynecol. 2014;210:179–93. 2. Childbirth Connection. Vaginal or cesarean Birth. What is at stake for Abbreviations women and babies. New York: Childbirth Connection; 2012. p. 52. CMDC: California Maternal Data Center; CNM: Certified nurse midwives; 3. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancy-related mortality in the family Medicine: Family medicine; family medicine-CS: Family medicine United States, 1987-1990. Obstet Gynecol. 1996;88:161–7. physicians with privileges to perform cesarean; HP2020: Healthy People 2020; 4. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancy-related LM: California licensed midwives; MFM: Maternal-fetal medicine; mortality in the United States, 1998 to 2005. Obstet Gynecol. 2010;116: NICU: Neonatal intensive care unit; NTSV: Nulliparous, term, singleton, vertex; 1302–9. OB: Obstetrician 5. Berg CJ, Chang J, Callaghan WM, Whitehead SJ. Pregnancy-related mortality in the United States, 1991-1997. Obstet Gynecol. 2003;101:289–96. Acknowledgements 6. Office of Disease Prevention and Health Promotion. Healthy people Jairo Chavez, supported by the Ambulatory Primary Care Innovations Group of 2020: maternal, infant, and child health. Washington (DC), 2016. the NorthShore University HealthSystem Research Institute assisted in the Retrieved from https://www.healthypeople.gov/2020/topics-objectives/ preparation of this manuscript. Suzanne Eidson-Ton, MD, supported by the topic/maternal-infant-and-child-health/objectives. Accessed 22 Nov 2016. University of California Davis Department of Family and Community Medicine, 7. National Quality Forum (NQF). National Voluntary Consensus Standards for contributed content expertise and assisted with pilot work for this study. perinatal care 2008: a consensus report. Washington, DC: NQF; 2009. Retrieved from https://www.qualityforum.org/Publications/2009/05/ Funding National_Voluntary_Consensus_Standards_for_Perinatal_Care_2008.aspx. This work was supported by grant number T32HS022236 from the Agency Accessed 22 Nov 2016. for Healthcare Research and Quality and grant number S-D936808-MCPAB 8. Main EK, Moore D, Farrell B, Schimmel LD, Altman RJ, Abrahams C, et al. Is from the University of California Davis Department of Family and Community there a useful cesarean birth measure? Assessment of the nulliparous term White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 9 of 9 singleton vertex cesarean birth rate as a tool for obstetric quality improvement, Am J Obstet Gynecol. 2006;194:1644-51. discussion 51-2 9. Kozhimannil KB, Law MR, Virnig BA. Cesarean delivery rates vary tenfold among US hospitals; reducing variation may address quality and cost issues. Health Aff (Millwood). 2013;32:527–35. 10. Caceres IA, Arcaya M, Declercq E, Belanoff CM, Janakiraman V, Cohen B, et al. Hospital differences in cesarean deliveries in Massachusetts (US) 2004- 2006: the case against case-mix artifact. PLoS One. 2013;8:e57817. 11. Metz TD, Allshouse AA, Gilbert SA, Doyle R, Tong A, Carey JC. Variation in primary cesarean delivery rates by individual physician within a single- hospital laborist model. Am J Obstet Gynecol. 2016;214:531 e1-6. 12. Main EK, Morton CH, Melsop K, Hopkins D, Giuliani G, Gould JB. Creating a public agenda for maternity safety and quality in cesarean delivery. Obstet Gynecol. 2012;120:1194–8. 13. Committee Opinion No. 687: approaches to limit intervention during labor and birth. Obstet Gynecol. 2017;129:e20–e8. 14. Olatunbosun OA, Edouard L, Pierson RA. Physicians' attitudes toward evidence based obstetric practice: a questionnaire survey. BMJ. 1998;316:365–6. 15. Reime B, Klein MC, Kelly A, Duxbury N, Saxell L, Liston R, et al. Do maternity care provider groups have different attitudes towards birth? BJOG. 2004;111:1388–93. 16. Klein MC, Kaczorowski J, Hall WA, Fraser W, Liston RM, Eftekhary S, et al. The attitudes of Canadian maternity care practitioners towards labour and birth: many differences but important similarities. J Obstet Gynaecol Can. 2009;31:827–40. 17. Burns LR, Geller SE, Wholey DR. The effect of physician factors on the cesarean section decision. Med Care. 1995;33:365–82. 18. Mitler LK, Rizzo JA, Horwitz SM. Physician gender and cesarean sections. J Clin Epidemiol. 2000;53:1030–5. 19. Tussing AD, Wojtowycz MA. The effect of physician characteristics on clinical behavior: cesarean section in New York state. Soc Sci Med. 1993;37:1251–60. 20. State of California, Office of Statewide Health Planning and Development, Healthcare Workforce Development Division. Medical service study areas. Sacramento: Office of Statewide Health Planning and Development; 2013. Retrieved from https://www.oshpd.ca.gov/HWDD/CalPCO.html#Maps-GIS. Accessed 21 Mar 2017. 21. Dillman DA, Smyth JD, Christian LM, Dillman DA. Internet, mail, and mixed-mode surveys : the tailored design method. 3rd ed. Hoboken: Wiley & Sons; 2009. 22. Cheng YW, Snowden JM, Handler S, Tager IB, Hubbard A, Caughey AB. Clinicians' practice environment is associated with a higher likelihood of recommending cesarean deliveries. J Matern Fetal Neonatal Med. 2014;27:1220–7. 23. Cheng YW, Snowden JM, Handler SJ, Tager IB, Hubbard AE, Caughey AB. Litigation in obstetrics: does defensive medicine contribute to increases in cesarean delivery? J Matern Fetal Neonatal Med. 2014;27:1668–75. 24. Chaillet N, Dumont A, Abrahamowicz M, Pasquier JC, Audibert F, Monnier P, et al. A cluster-randomized trial to reduce cesarean delivery rates in Quebec. N Engl J Med. 2015;372:1710–21. 25. Asch DA, Nicholson S, Srinivas S, Herrin J, Epstein AJ. Evaluating obstetrical residency programs using patient outcomes. JAMA. 2009;302:1277–83. 26. Bardos J, Loudon H, Rekawek P, Friedman F, Brodman M, Fox NS. Association between senior obstetrician supervision of resident deliveries and mode of delivery. Obstet Gynecol. 2017;129:486–90. 27. King TL, Laros RK, Parer JT. Interprofessional collaborative practice in obstetrics and midwifery. Obstet Gynecol Clin N Am. 2012;39:411–22. 28. Dan O, Hochner-Celnikier D, Solnica A, Loewenstein Y. Association of Catastrophic Neonatal Outcomes with Increased Rate of subsequent cesarean deliveries. Obstet Gynecol. 2017;129:671–5. 29. Colegrave N, Ruxton G. Confidence intervals are a more useful complement to nonsignificant tests than are power calculations. Behav Ecol. 2003;14:446–7. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png BMC Pregnancy and Childbirth Springer Journals

Do provider birth attitudes influence cesarean delivery rate: a cross-sectional study

Free
9 pages
Loading next page...
 
/lp/springer_journal/do-provider-birth-attitudes-influence-cesarean-delivery-rate-a-cross-BBqVQd880J
Publisher
BioMed Central
Copyright
Copyright © 2018 by The Author(s).
Subject
Medicine & Public Health; Reproductive Medicine; Maternal and Child Health; Gynecology
eISSN
1471-2393
D.O.I.
10.1186/s12884-018-1756-7
Publisher site
See Article on Publisher Site

Abstract

Background: When used judiciously, cesarean sections can save lives; but in the United States, prior research indicates that cesarean birth rates have risen beyond the threshold to help women and infants and become a contributor to increased maternal mortality and rising healthcare costs. Healthy People 2020 has set the goal for nulliparous, term, singleton, vertex (NTSV) cesarean birth rate at no more than 23.9% of births. Currently, cesarean rates vary from 6% to 69% in US hospitals, unexplained by clinical or demographic factors. This wide variation in cesarean use is also seen among individual providers of intrapartum care. Previous research of birth attitudes found providers of intrapartum care hold widely differing views, which may be a key underlying factor influencing practice variation; however, further study is needed to determine if differences in attitudes are associated with differences in clinical outcomes. The purpose of this study was to estimate the association between individual provider attitudes towards birth and their low-risk primary cesarean rate. Methods: Four hundred providers were drawn from a stratified random sample of all California providers of intrapartum care in 2013 and surveyed for their attitudes towards various aspects of labor and birth. Providers’ NTSV cesarean birth rates were obtained for 2013 and 2014. Covariates included gender, years of experience, practice location, and primary hospital’s NTSV cesarean rate. We used adjusted multivariate Poisson regression to compare cesarean rates and linear regression to compare attitude scores of providers meeting versus not meeting the Healthy People 2020 (HP2020) goal. Results: Two hundred nine total participants (obstetricians, family physicians, and midwives) completed surveys, of which 109 perform cesareans. Providers’ NTSV cesarean rate was significantly associated with their composite attitudes score [IRR for each one-point increase 1.21 (95% CI 1.002–1.45)]. Physicians meeting the HP2020 goal held attitudes which were significantly more favorable towards vaginal birth: mean 2.70 (95% CI 2.58–2.83) versus 2.91 (95% CI 2.82–3.00), p < 0.01. Conclusions: Provider attitudinal differences are associated with NTSV cesarean rates. Those meeting the HP2020 goal hold attitudes more favorable towards vaginal birth. These findings may present a modifiable target for quality improvement initiatives to decrease low risk primary cesareans. Keywords: Cesarean section, Provider attitudes, Quality improvement, Primary cesarean, Culture * Correspondence: EWhiteVanGompel@northshore.org Department of Family Medicine, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, Evanston, IL 60201, USA Department of Obstetrics and Gynecology, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, Evanston, IL 60201, USA Full list of author information is available at the end of the article © The Author(s). 2018 Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated. White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 2 of 9 Background service of the California Maternal Quality Care Collab- Cesarean sections, when used judiciously, save lives; orative. Baseline demographics and survey scores were however, research indicates that the use of cesareans in analyzed using SAS Version 9.4. Poisson regressions were the United States has risen well above the level of neces- performed using Stata 12.1. This study was approved sity and has become a contributor to maternal morbidity under expedited review by University of California Davis and mortality [1–5]. Women who are nulliparous, at full Internal Review Board. Participants received the consent term, with a singleton pregnancy in vertex presentation document, but signed consent was waived by the IRB. (NTSV) have been established as a standard population and used as a target group for reducing the cesarean Setting birth rate [6–8]. Healthy People 2020 has set the goal California has approximately 3500 providers that are for NTSV cesarean rate at no more than 23.9% of births listed as delivering clinicians in a given year; these [6]. Currently, cesarean rates vary from 6% to 69% in US include obstetricians (OB), Maternal-Fetal Medicine hospitals [9]. Substantial variation persists after adjust- specialists, family medicine physicians, Certified Nurse ment for hospital demographics, referral categories, or Midwives (CNM), and California Licensed Midwives teaching status, and after adjustment for patient clinical (LM). Practice settings are diverse, and include rural, or sociodemographic factors [9, 10]. urban, and frontier geographic locations, teaching hospi- Wide variation is also seen between providers, even tals, community hospitals, and tertiary care centers. Pro- those practicing within the same hospital and utilizing a vider mix at each hospital may include only obstetricians laborist model [11]. Over 90% of the variation in the or family physicians, or may include Maternal-Fetal NTSV cesarean rate is due to two indications: “fetal in- Medicine specialists or midwives. The CMDC combines tolerance of labor” and “failure to progress”; indications existing datasets, including monthly discharge and requiring subjective decision-making by the intrapartum clinical data, birth certificate data, and semi-annual patient provider [12]. The American Congress of Obstetricians dischargedatafromthe California Office of Statewide and Gynecologists’ Committee Opinion on limiting Health Planning and Development, to create hospital and intervention during labor highlights the labor manage- physician-level quality improvement metrics. We obtained ment techniques that, despite prevailing evidence, vary a stratified random sample from 2013 - the latest available significantly among providers [13]. Evidence is filtered year of complete data. At the time of final data collection, through the lens of a provider’s experiences and 2014 complete data became available, thus we included attitudes [14]; yet how these attitudes affect clinical out- provider and hospital metrics from both 2013 and 2014 comes has not been evaluated. combined. Our primary sample size consideration was Previous research found differences in birth attitudes based on the six domain scores’ ability to account for vari- between providers of different disciplines (obstetrics, ation in NTSV cesarean delivery rate. Using pilot data, we family medicine, midwifery) [15, 16]; yet did not exam- determined that a sample size of 116 patients would ine clinical outcomes. The objective of this study was to provide 80% power to detect an 11.1% partial R-square. To examine the association of individual provider NTSV account for intermittent missing variables, we targeted a cesarean rates with their attitudes towards birth. In con- sample of 130 survey participants. trast to earlier studies examining physician demographic Inclusion criteria for providers were: having been listed factors [17–19], this study sought to evaluate a poten- on a birth certificate as the delivering clinician in the tially modifiable personal attribute that may contribute year 2013, belonging to one of the key study disciplines to cesarean overuse. (maternal-fetal medicine, OB, family medicine, CNM, LM), and having performed at least 10 deliveries per Aim year. Providers with a license address outside the state of Our primary aim was to assess the association between California and those without an identifiable discipline providers’ birth attitudes and their NTSV cesarean rates. based either on license number prefix or NPI taxonomy code were excluded from sampling. Methods Based on prior research pointing to important influ- Design ences on provider practice [17–19], stratification was A stratified random sample of providers listed as deliver- performed based on three variables: provider discipline ing clinicians on California birth certificates in 2013 (maternal-fetal medicine, OB, family medicine, CNM, were surveyed with a previously validated survey instru- LM), geographic location as defined by the federally des- ment of provider birth attitudes [16]. We linked pro- ignated Medical Service Study Areas (rural = population viders’ scores on this survey with their individual NTSV density < 250 persons/square mile; frontier = population cesarean rates for the years of 2013 and 2014, as calcu- density < 11 persons/square mile; urban = anything not lated by the California Maternal Data Center (CMDC), a rural or frontier) [20], and years in practice (< 5 years, White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 3 of 9 5–15 years, 16–25 years, > 25 years). The stratified ran- a provider that qualify as primary low-risk cesareans. This dom sample was drawn in two rounds of 200, for a total criteria includes all nulliparous, term (> 37 weeks gestation), of 400 sampled providers. To ensure adequate power for singleton gestations in vertex presentation at delivery. These analysis, oversampling was performed on provider are termed NTSV births for nulliparous, term, singleton, groups other than obstetricians (who perform the vast and vertex, and remain the primary target for quality im- majority of all births in California). Researchers were provement initiatives to decrease cesarean section overuse blinded to individual providers’ cesarean rates during in the United States [6–8]. We obtained both the total num- sample selection. ber of NTSV births providers attended in 2013–2014, and the total number of those births that were delivered via Attitudes survey cesarean from the CMDC database. The birth attitudes survey, previously validated by Klein Not all providers included in the survey sample had privi- and colleagues, included 9 different domains assessing pro- leges to perform cesareans. In order to provide a complete vider attitudes towards different aspects of labor and birth picture of the spectrum of attitudes held by all providers [16]. Six of these domains, comprising 31 Likert-style items, practicing as independent clinicians in California, we chose were chosen as thosemostlikelytohaveaneffect on the to include all those listed as delivering clinicians on at least targeted outcome – low-risk primary cesareans. They in- 10 deliveries per study year. As it would not be appropriate cluded (renumbered for this study): Domain 1: attitudes re- to assess provider-level cesarean rates in providers who garding use of electronic fetal monitoring (Cronbach alpha cannot perform cesareans, we excluded these providers [α] = 0.704), Domain 2: factors that increase cesarean rates from theNTSVpredictivemodel.BothCNMsand LMsdo (α = 0.810), Domain 3: fears of birth mode by respondents not perform cesareans under any circumstances. A majority or their partners/spouses (α=0.929),Domain4:factorsthat of family medicine physicians who do prenatal and intra- decrease cesarean rates (α = 0.819), Domain 5: maternal partum care do not perform cesareans; however, there are choice and mothers’ roles in birth (α = 0.646), and Domain family medicine physicians who have done additional train- 6: safety by mode or place of birth (α = 0.748). The compos- ing to qualify to perform cesareans. There is no central ite scale combined the individual domain scores. Coding of database that tracks these privileges, thus this was ascer- domains 4 and 5 was reversed for directional consistency to tained by anonymously calling individual family medicine create a total mean score. Lower scores on the composite physicians’ offices. scale indicate attitudes more favorable toward vaginal birth, while higher scores indicate attitudes that favor cesarean Covariates birth. Providers were also asked to give their discipline and Provider demographic data, including discipline (mater- years of experience. Each provider was assigned a random nal-fetal medicine, Average Risk OB, family medicine, identifier, so that survey responses were not associated with CNM, LM), gender, years since graduation from medical provider name. school (< 5 years, 5–15 years, 16–25 years, > 25 years), We used best practices for achieving maximal survey practice geographic location (rural/urban), and hospital- response as detailed by Dillman et al. [21] Survey data level demographics including hospital-level NTSV rate, collection began in October 2015 and ended in April neonatal intensive care unit level, and percent Medicaid 2016. We sent the attitudes survey via postal mail and were obtained from the CMDC database. Some providers included a web address for optional completion online. practiced at multiple hospitals; however, the hospital The initial mailing included the 4-page survey, a cover where they had their greatest number of NTSV births was letter, informed consent document, a self-addressed assigned as their primary hospital of delivery. stamped envelope, and an incentive of a $10 Starbucks gift card. Providers could opt to provide their email Statistical methods addresses to be entered into a drawing at the conclusion Due to the large range of providers’ total NTSV birth vol- of the study for an iPad of approximately $400 value at ume, we used multivariate Poisson regression to maximize the conclusion of the study. One to two weeks after the precision of our estimation model, using counts of NTSV initial mailing, we sent a postcard reminder to complete cesarean deliveries per provider as the dependent variable the survey. Between one and 2 months after the initial and total deliveries per provider as an “exposure” variable mailing, all non-responders were sent a second complete whose log-transformed value is included as an offset term, packet including cover letter, consent, survey document, to account for between-provider variation. We used and self-addressed stamped envelope. robust standard error estimators to protect against model misspecification. Simple regression with predicted margins Provider-level cesarean rate was used to compare attitudinal scores between discipline The CMDC database uses the standard NQF-endorsed algo- groups. Attitude domain and composite means were cal- rithm [4] to calculate the total number of births attended by culated from all non-missing items. For the dichotomous White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 4 of 9 Healthy People 2020 provider comparison, we compared had higher annual birth volumes than non-respondents’ providers’ attitudes based on whether or not their NTSV hospitals (mean = 5,203.9 versus 2,666.8, p < 0.001). Hos- cesarean rates met the Healthy People 2020 (HP2020) goal pital NTSV rates were slightly lower for respondents com- of less than 23.9% [6]. We used multiple linear regression, pared with non-respondents (25.3% versus 26.8%, p < 0.01). adjusting for provider gender, practice geography, and (Table 1). experience level to compare predicted means for each Attitudes varied significantly according to provider group. In order to adjust for the provider’shospital discipline (Fig. 1). Each domain displayed a spectrum of cultural environment, we adjusted each regression model provider attitudes towards birth, with midwives on one for the primary hospital’s NTSV rate exclusive of pro- end of the spectrum and OBs on the other (Fig. 2). Fam- vider’scontributionto this rate. ily medicine physicians either fell between the two or held attitudes more similar to midwives. Maternal-fetal Results medicine physicians held attitudes most consistent with We received 209 completed surveys, a total response rate average risk OBs, but tended to express attitudes that of 52.3%, including 22 maternal-fetal medicine, 101 OB, 53 fell between average risk OBs and family medicine physi- family medicine, 16 CNM, and 17 LM responding. There cians. The OB group had widest variation in composite was a higher response rate for midwives, 97% of who were attitude scores (range = 1.37 to 4.33). female, but no difference in disciplinary distribution or Of the 53 responding family medicine physicians, only gender of physician respondents versus non-respondents. 8 had privileges to perform cesareans, though 22 had Experience level, practice geographic location, and primary “first assist” privileges. Family medicine physicians with hospital nursery level did not vary between respondents cesarean privileges (family medicine-CS) were identical versus non-respondents. Respondents’ primary hospitals to family medicine physicians without privileges on the Table 1 Demographic and Practice Characteristics of Responders and Non-responders Responders (n) Percent Non-Responders (n) Percent p-value N 209 191 Discipline 0.35 MFM 22 10.5 19 9.9 OB 101 48.3 120 62.8 Family Medicine 53 25.4 46 24.1 CNM 16 7.7 3 2.1 LM 17 8.1 3 1.6 Gender 0.29 Female 132 63.2 102 53.4 Male 77 36.8 89 46.6 Experience 0.49 < 5 years 35 18.3 28 13.9 5–15 years 58 30.4 70 34.7 16–25 years 46 24.1 43 21.3 > 25 years 52 27.2 61 30.2 Hospital NICU Level 0.64 Basic Nursery 50 23.9 57 29.8 Community Nursery 82 39.2 73 38.2 Intermediate Nursery 36 17.2 33 17.3 Regional Nursery 20 9.6 25 13.1 Continuous Variables Mean SD Mean SD p-value Hospital Birth Volume 5203.9 3835.9 2666.8 2012.8 < 0.0001 Hospital NTSV CS Rate 25.30 4.80 26.80 6.10 < 0.01 Abbreviations: MFM maternal fetal medicine, OB obstetricians, Family Medicine family medicine physicians, CNM certified nurse midwives, LM licensed midwives, NICU neonatal intensive care unit, NTSV CS nulliparous, term, singleton, vertex cesarean section P-value includes physicians only. Midwives were all female except for one participant and had exceptionally high response rate Calculated for provider’s primary delivering hospital site White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 5 of 9 Fig. 1 Mean Provider Attitudes Scores on the Composite Scale by Provider Discipline. Figure 1 is a boxplot of providers’ mean attitude scores on the composite attitudes scale categorized by provider membership in a training discipline. Abbreviations: MFM: maternal fetal medicine; OB: obstetricians; family medicine: family medicine physicians; CNM: certified nurse midwives; LM: licensed midwives composite scale (mean 2.31 with SD 0.16, and 2.31 with providers meeting this goal and 65% of providers SD 0.35, respectively, p = 0.98). For the most part, family whose rate exceeded the goal. Providers meeting the medicine-CS physicians expressed attitudes on the individ- HP2020 goal held attitudes more favorable toward ual domains that were slightly less favorable towards vagi- vaginal birth compared with those over the HP2020 nal delivery than family medicine physicians; however, goal [adjusted mean 2.70 (95% CI 2.58–2.83) versus family medicine-CS tended to endorse less fear of vaginal 2.91 (95% CI 2.82–3.00), p < 0.01]. delivery than their family medicine counterparts, though this was not statistically significant (mean 1.19 ± 0.30 versus Discussion 1.40 ± 0.74, respectively, p = 0.43). The database attributed We found a significant association between providers’ cesarean deliveries to 14 family medicine physicians with- attitudes and beliefs about birth to their own NTSV out any cesarean privileges and 10 family medicine physi- cesarean rate. This study was consistent with the find- cians with first assist privileges. ings of the original Canadian survey validation study The analyses of association of provider attitudes with [16], finding that California providers’ attitudes towards NTSV cesarean rates included only providers with cesarean birth were primarily divided along disciplinary lines. privileges and at least 20 NTSV deliveries within the 2 year One of the most surprising findings in the disciplinary study period: 11 maternal-fetal medicine, 91 OB, and 7 analysis was the wide range of obstetricians’ attitudes, family medicine physicians. The composite attitudes scale spanning views more pro-cesarean than maternal-fetal scores for these providers ranged from 1.73 to 4.33 (mean 2. medicine providers all the way to those consistent with 83, standard deviation 0.48). Adjusted incidence rate ratios midwives. This large variation within a single discipline for domain and composite scale from the Poisson multiple suggests that there may be key acculturation differences regression are shown in Table 2; this model adjusted for in obstetric training and practice environment, where a provider gender, practice geography, experience level, and provider’s attitudes and beliefs are influenced. primary hospital’s NTSV rate (exclusive of provider’scontri- We took prior work an important step further by con- bution to this rate). The composite attitudes scale had an necting provider attitudes to their own measured clinical incident rate ratio of 1.21 (95% CI 1.002–1.45, p = 0.048), outcomes. For providers with cesarean privileges, those indicating that, for every 1 point increase in a provider’s meeting the Healthy People 2020 goal held attitudes that score on the composite scale, their NTSV cesarean rate were more favorable toward vaginal birth than those not decreased relatively by 21%. For individual domains, meeting this goal. In our regression analysis, as attitudes attitudes towards the perceived benefits of cesarean and fear became more favorable toward cesarean and less favor- of vaginal birth approached significance [respectively, 1.07 able toward vaginal birth, a provider’s NTSV cesarean (95% CI 0.99–1.17) and 1.08 (95% CI 0.99–1.17)]. (Table 2). rate increased proportionately. Additionally, we used a When dichotomizing providers by the Healthy People novel adjustor to account for local hospital culture – the 2020 NTSV cesarean goal cut-off of 23.9%, our sample primary hospital’s NTSV cesarean rate less the provider’s reflected the overall California distribution, with 35% of contribution to that rate. This adjustment highlights the White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 6 of 9 Fig. 2 a-f Boxplots of Provider Attitudes Scores on the Individual Domain Scales by Provider Discipline. Figure 2 a-f includes individual boxplots of providers’ mean attitude scores for each individual attitudinal domain categorized by provider membership in a training discipline. Abbreviations: MFM: maternal fetal medicine; OB: obstetricians; family medicine: family medicine physicians; CNM: certified nurse midwives; LM: licensed midwives White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 7 of 9 a b Table 2 Adjusted Associations between Attitude Scores and Provider NTSV CS Rate Domain Incidence Rate Ratio 95% CI p-value Fetal Monitoring 1.01 0.94–1.10 0.71 Benefits of CS 1.07 0.99–1.17 0.10 Fear of Vaginal Birth 1.08 0.99–1.17 0.09 Measures to Decrease CS 0.91 0.77–1.08 0.28 Maternal Role in Birth 0.93 0.85–1.02 0.14 Safety by Place or Mode of Birth 1.12 0.97–1.30 0.14 Composite Scale 1.21 1.002–1.45 0.048 Abbreviations: NTSV CS nulliparous, term, singleton, vertex cesarean section, CS cesarean section, CI confidence interval Results are from individual Poisson regression models of the NTSV cesarean outcome, with one model for each row, with that variable the focal predictor and with additional covariates used to adjust for gender, experience level, geographic location of practice, primary hospital’s NTSV cesarean rate calculated without the individual provider’s contribution Only includes providers with confirmed privileges to perform cesarean sections who had at least 20 NTSV births over the two-year study period of 2013–2014 Higher scores indicate attitudes more favorable toward cesarean section except for the two reverse-coded scales (as below) Higher scores indicate attitudes more favorable toward vaginal birth on these scales impact of individual provider attitudes on top of that of program significantly decreased their primary cesarean local hospital norms and practices. These findings sug- rate by providing senior obstetric supervision of resi- gest that a provider’s underlying attitudes, values, and dents on labor and delivery, highlighting the impact of beliefs play an important role in intrapartum decisions preceptor experience level on trainees [26]. The impact that ultimately affect birth outcomes. of integrating midwives into traditional obstetric training Our results are consistent with, and provide a possible has been posited but not yet rigorously tested against underlying mechanism for, other studies that have found clinical outcomes [27]. Finally, ongoing training and provider differences such as demographics, litigation support after experiencing a traumatic delivery event history, and practice environment are associated with may mitigate some of the fear attitudes associated with provider self-report personal thresholds to perform increased cesarean rate, which appear to impact entire cesareans [22, 23]. By independently measuring attitudes hospital units and not just the providers involved. This while protecting the confidentiality of providers and was described recently in a study of unplanned hospital using administrative data to associate attitudes with an cesarean rates, which increased and stayed elevated for unbiased measure of the individual provider-level 4 weeks after any catastrophic neonatal outcome within cesarean rate, our study minimized social desirability that hospital [28]. bias, which can confound self-report data. Of note, the actual magnitude of the point estimate of Limitations the association between composite attitudinal score and This cross-sectional study cannot draw conclusions NTSV cesarean rate is quite large. For every 1 point regarding causality or time course between independent increase in agreement with attitudes favoring cesarean, and dependent variables. Additionally, our a priori sam- NTSV cesarean rates increased by a relative 21%. For a ple size justification assumed a linear regression analysis, provider with a baseline rate of 25%, this would translate but we found that a Poisson regression analysis was bet- into an absolute change of 5.25%. In comparison, the ter suited to the outcome distribution. Thus, in order to QUARISMA trial found changes to the absolute assess the adequacy of the realized sample size for the cesarean rate of 0.7–2.3% resulting from an audit and reported effect sizes, one should consider the range of feedback mechanism in conjunction with hospital-based values included within the 95% CI [29]. Where the null best practices implementation [24]. Targeting attitudinal value was included along with values that would be change, which is a “bottom-up” method, in conjunction clinically meaningful, one could conclude that the effect with quality improvement, or “top-down”, efforts may is ambiguous and would require a more precise estimate enhance the impact of interventions. Research is needed in future work. For example, we would assert that a 15% to identify effective interventions to enhance evidence- relative increase in the NTSV cesarean rate would be based attitudes towards vaginal birth. particularly clinically meaningful. When examining the Training may offer a promising target for influencing 95% CI for the effect size of the individual attitude items, attitudes that favor vaginal birth. One study found that we find that all of them included the null value (i.e. were providers who trained in hospitals with lower obstetric not statistically significant) but three of them extended complication rates continued to have lower complication beyond an IRR of 1.15, suggesting that the results for rates once in practice [25]. Most recently, one hospital those items are ambiguous and could warrant a study with White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 8 of 9 a larger sample size. For the composite attitude meas- Medicine Jump Start Fund. Neither funding body had any role in the design, collection, analysis, interpretation of data, or in writing the manuscript. ure, we found statistically significant and clinically significant effects. Availability of data and materials Finally, the administrative data used to assess provider- The data that support the findings of this study are available from the California Maternal Data Center, but restrictions apply to the availability of these data, level cesarean rate was restricted to the provider listed on which were used under license for the current study, and so are not publicly the birth certificate and subject to errors of attribution. For available. Data are however available from the authors upon reasonable request example, the provider listed as “Delivering Clinician” on the and with permission of the California Maternal Data Center. Birth Certificate is usually but not always the person that Authors’ contributions managed the majority of a patient’slabor or made the EWV and JM designed the study with statistical consultation from DT and decision to go to cesarean. Unfortunately, this method content area expertise from EM. EM created the provider data file for sampling and supplied outcomes data for participants. EWV and DT analyzed discounts information about providers who do not carry and interpreted the attitudinal and outcomes data. EWV wrote the majority cesarean privileges but may play a major role in intrapar- of the manuscript with major contributions from all authors. All authors read tum care, and may be key decision-makers along the route and approved the final manuscript. that ends in either cesarean or vaginal delivery. For Ethics approval and consent to participate example, a provider may decide to admit a patient prior to This study was approved under expedited review by the University of the onset of active labor or use continuous electronic fetal California Davis Internal Review Board. Participants received the consent document, but signed consent was waived by the IRB. monitoring despite a patient’s low-risk status, both of which increase the likelihood of that patient requiring a cesarean, Competing interests yet the cesarean birth would be attributed to the clinician The authors declare that they have no competing interests. who performed the surgery itself and not the provider man- aging the labor. Further studies are needed to assess the Publisher’sNote Springer Nature remains neutral with regard to jurisdictional claims in interplay of personnel on labor and delivery wards, how key published maps and institutional affiliations. decisions are made and birth outcomes attributed. Author details Department of Family Medicine, The University of Chicago, Pritzker School Conclusions of Medicine, NorthShore University HealthSystem Research Institute, 1001 This is the first study we are aware of linking provider attitu- 2 University Place, Evanston, IL 60201, USA. Department of Obstetrics and dinal differences to differences in measured birth outcomes. Gynecology, The University of Chicago, Pritzker School of Medicine, NorthShore University HealthSystem Research Institute, 1001 University Place, We found that the more providers’ birth attitudes favored Evanston, IL 60201, USA. California Maternal Quality Care Collaborative, cesarean, the higher their NTSV cesarean rate. In contrast Stanford University, Stanford Medical School Office Building, 1265 Welch to earlier studies that have focused on physician demograph- Road, MS 5415, Stanford, CA 94305, USA. Center for Healthcare Policy and Research and Department of Pediatrics, University of California Davis School ics or litigation history, which arerarelymodifiablewithout of Medicine, 2103 Stockton Blvd, Sacramento, CA 95817, USA. Center for major policy changes, we sought to evaluate the influence of Healthcare Policy and Research and Department of Family and Community a modifiable attribute – aprovider’s personal attitudes and Medicine, University of California Davis School of Medicine, 2103 Stockton Blvd, Sacramento, CA 95817, USA. beliefs. These findings suggest further scrutiny is needed of how future providers are acculturated during training and Received: 28 July 2017 Accepted: 19 April 2018 while in practice, and how practice groups, hospital units, and inter-professional interactions may modify these References attitudes to ultimately improve quality of care and health 1. Caughey AB, Cahill AG, Guise JM, Rouse DJ. Safe prevention of the primary outcomes. cesarean delivery. Am J Obstet Gynecol. 2014;210:179–93. 2. Childbirth Connection. Vaginal or cesarean Birth. What is at stake for Abbreviations women and babies. New York: Childbirth Connection; 2012. p. 52. CMDC: California Maternal Data Center; CNM: Certified nurse midwives; 3. Berg CJ, Atrash HK, Koonin LM, Tucker M. Pregnancy-related mortality in the family Medicine: Family medicine; family medicine-CS: Family medicine United States, 1987-1990. Obstet Gynecol. 1996;88:161–7. physicians with privileges to perform cesarean; HP2020: Healthy People 2020; 4. Berg CJ, Callaghan WM, Syverson C, Henderson Z. Pregnancy-related LM: California licensed midwives; MFM: Maternal-fetal medicine; mortality in the United States, 1998 to 2005. Obstet Gynecol. 2010;116: NICU: Neonatal intensive care unit; NTSV: Nulliparous, term, singleton, vertex; 1302–9. OB: Obstetrician 5. Berg CJ, Chang J, Callaghan WM, Whitehead SJ. Pregnancy-related mortality in the United States, 1991-1997. Obstet Gynecol. 2003;101:289–96. Acknowledgements 6. Office of Disease Prevention and Health Promotion. Healthy people Jairo Chavez, supported by the Ambulatory Primary Care Innovations Group of 2020: maternal, infant, and child health. Washington (DC), 2016. the NorthShore University HealthSystem Research Institute assisted in the Retrieved from https://www.healthypeople.gov/2020/topics-objectives/ preparation of this manuscript. Suzanne Eidson-Ton, MD, supported by the topic/maternal-infant-and-child-health/objectives. Accessed 22 Nov 2016. University of California Davis Department of Family and Community Medicine, 7. National Quality Forum (NQF). National Voluntary Consensus Standards for contributed content expertise and assisted with pilot work for this study. perinatal care 2008: a consensus report. Washington, DC: NQF; 2009. Retrieved from https://www.qualityforum.org/Publications/2009/05/ Funding National_Voluntary_Consensus_Standards_for_Perinatal_Care_2008.aspx. This work was supported by grant number T32HS022236 from the Agency Accessed 22 Nov 2016. for Healthcare Research and Quality and grant number S-D936808-MCPAB 8. Main EK, Moore D, Farrell B, Schimmel LD, Altman RJ, Abrahams C, et al. Is from the University of California Davis Department of Family and Community there a useful cesarean birth measure? Assessment of the nulliparous term White VanGompel et al. BMC Pregnancy and Childbirth (2018) 18:184 Page 9 of 9 singleton vertex cesarean birth rate as a tool for obstetric quality improvement, Am J Obstet Gynecol. 2006;194:1644-51. discussion 51-2 9. Kozhimannil KB, Law MR, Virnig BA. Cesarean delivery rates vary tenfold among US hospitals; reducing variation may address quality and cost issues. Health Aff (Millwood). 2013;32:527–35. 10. Caceres IA, Arcaya M, Declercq E, Belanoff CM, Janakiraman V, Cohen B, et al. Hospital differences in cesarean deliveries in Massachusetts (US) 2004- 2006: the case against case-mix artifact. PLoS One. 2013;8:e57817. 11. Metz TD, Allshouse AA, Gilbert SA, Doyle R, Tong A, Carey JC. Variation in primary cesarean delivery rates by individual physician within a single- hospital laborist model. Am J Obstet Gynecol. 2016;214:531 e1-6. 12. Main EK, Morton CH, Melsop K, Hopkins D, Giuliani G, Gould JB. Creating a public agenda for maternity safety and quality in cesarean delivery. Obstet Gynecol. 2012;120:1194–8. 13. Committee Opinion No. 687: approaches to limit intervention during labor and birth. Obstet Gynecol. 2017;129:e20–e8. 14. Olatunbosun OA, Edouard L, Pierson RA. Physicians' attitudes toward evidence based obstetric practice: a questionnaire survey. BMJ. 1998;316:365–6. 15. Reime B, Klein MC, Kelly A, Duxbury N, Saxell L, Liston R, et al. Do maternity care provider groups have different attitudes towards birth? BJOG. 2004;111:1388–93. 16. Klein MC, Kaczorowski J, Hall WA, Fraser W, Liston RM, Eftekhary S, et al. The attitudes of Canadian maternity care practitioners towards labour and birth: many differences but important similarities. J Obstet Gynaecol Can. 2009;31:827–40. 17. Burns LR, Geller SE, Wholey DR. The effect of physician factors on the cesarean section decision. Med Care. 1995;33:365–82. 18. Mitler LK, Rizzo JA, Horwitz SM. Physician gender and cesarean sections. J Clin Epidemiol. 2000;53:1030–5. 19. Tussing AD, Wojtowycz MA. The effect of physician characteristics on clinical behavior: cesarean section in New York state. Soc Sci Med. 1993;37:1251–60. 20. State of California, Office of Statewide Health Planning and Development, Healthcare Workforce Development Division. Medical service study areas. Sacramento: Office of Statewide Health Planning and Development; 2013. Retrieved from https://www.oshpd.ca.gov/HWDD/CalPCO.html#Maps-GIS. Accessed 21 Mar 2017. 21. Dillman DA, Smyth JD, Christian LM, Dillman DA. Internet, mail, and mixed-mode surveys : the tailored design method. 3rd ed. Hoboken: Wiley & Sons; 2009. 22. Cheng YW, Snowden JM, Handler S, Tager IB, Hubbard A, Caughey AB. Clinicians' practice environment is associated with a higher likelihood of recommending cesarean deliveries. J Matern Fetal Neonatal Med. 2014;27:1220–7. 23. Cheng YW, Snowden JM, Handler SJ, Tager IB, Hubbard AE, Caughey AB. Litigation in obstetrics: does defensive medicine contribute to increases in cesarean delivery? J Matern Fetal Neonatal Med. 2014;27:1668–75. 24. Chaillet N, Dumont A, Abrahamowicz M, Pasquier JC, Audibert F, Monnier P, et al. A cluster-randomized trial to reduce cesarean delivery rates in Quebec. N Engl J Med. 2015;372:1710–21. 25. Asch DA, Nicholson S, Srinivas S, Herrin J, Epstein AJ. Evaluating obstetrical residency programs using patient outcomes. JAMA. 2009;302:1277–83. 26. Bardos J, Loudon H, Rekawek P, Friedman F, Brodman M, Fox NS. Association between senior obstetrician supervision of resident deliveries and mode of delivery. Obstet Gynecol. 2017;129:486–90. 27. King TL, Laros RK, Parer JT. Interprofessional collaborative practice in obstetrics and midwifery. Obstet Gynecol Clin N Am. 2012;39:411–22. 28. Dan O, Hochner-Celnikier D, Solnica A, Loewenstein Y. Association of Catastrophic Neonatal Outcomes with Increased Rate of subsequent cesarean deliveries. Obstet Gynecol. 2017;129:671–5. 29. Colegrave N, Ruxton G. Confidence intervals are a more useful complement to nonsignificant tests than are power calculations. Behav Ecol. 2003;14:446–7.

Journal

BMC Pregnancy and ChildbirthSpringer Journals

Published: May 29, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off