J. Perinat. Med. 2015; 43(1): 37–41
Sanjay Manohar, Christopher F.G. Woods and Stephen W. Lindow*
Individual consultant practice does not affect the overall intervention rate: a 6-year study Background
Abstract Background: Differences exist in obstetric intervention rates between hospitals but it is not known if the individual consultant governs the decision to intervene or whether intervention is a product of agreed protocols and working practices. The purpose of this study is to analyse the differences in obstetric intervention rates amongst individual consultants working in a large maternity unit. Methods: Each consultant was responsible for all deliveries occurring in successive 24-h periods. Over a 6-year period all deliveries resulting from a spontaneous onset of labour were matched to the consultant in charge at the time of the delivery and analysed. Results: There were no differences seen in normal delivery rates (χ2 = 4.478, P = 0.812) and vacuum (χ2 = 12.232, P = 0.141) rates for the consultants. Significant differences were found in both forceps rate (χ2 = 21.462, P = 0.006) and caesarean rate (χ2 = 24.535, P = 0.002) between consultants. When the forceps rate was combined with vacuum rate there were no significant differences. Conclusions: Within the hospital, individual consultants demonstrated no significant variations in overall intervention rates. However, when intervention occurred, different consultants showed preferences for forceps and caesarean section. Keywords: Caesarean; labour management; specialist obstetrician. DOI 10.1515/jpm-2014-0021 Received January 20, 2014. Accepted April 4, 2014. Previously published online May 8, 2014.
*Corresponding author: Stephen W. Lindow, Senior Lecturer in Perinatology, Women and Children’s Hospital, Hull and East Yorkshire Hospitals, NHS Trust, Kingston upon Hull, East Riding of Yorkshire, UK, Phone: +01482 875875, Fax: +01482 382781, E-mail: [email protected] Sanjay Manohar and Christopher F.G. Woods: Hull York Medical School, Hull Royal Infirmary, Hull, UK
The UK is one of the safest places to give birth, with a maternal mortality rate of 11.39 per 100,000 m aternities [6]. Current obstetric practice in the UK is guided by the Royal College of Obstetricians and Gynaecologists (RCOG) which has outlined standards for clinical practice and issued guidelines for the management of specific conditions [10]. Guidelines are, however, open to interpretation by clinicians who frequently have their own opinions on preferred management options [2]. Labour ward management involves the interaction of senior and junior obstetricians, anaesthetists and midwives on a regular basis [7]. The decisions relating to obstetric intervention could be governed by an individual consultant or by the generally agreed protocols operative in a unit. Previous work has suggested that there are indeed differences in the way patients are treated by health professionals within specialities for similar medical problems [13]. The purpose of this study is to investigate whether obstetric consultants working in the same hospital managed labour differently when compared to one another, in terms of overall intervention rates.
Methods This study has used existing data from the Obstetric department in the Hull and East Yorkshire Women and Children’s hospital. The obstetric department delivers over 5500 live babies every year and consists of 11 consultants who are in turn responsible for 24-h shifts. Consultants are physically present in the labour ward for 8–12 h per day (4 h minimum at weekends) and on call from home the rest of the time. A consultant works with a registrar and senior house officer who are present for 24 h on a shift pattern of work. All caesarean sections must be authorised by a consultant but forceps and vacuum extractions can be authorised and performed by a registrar. The existing data were recorded immediately after each individual delivery by the midwife who was responsible for giving care to the patient. In order to have a fair comparison between consultants, patients who had labour induction of any form, elective caesarean section, twins and stillbirths were removed, leaving only spontaneous onset of labour with single live births for analysis. Each patient’s delivery times were then cross-matched with the corresponding
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
38 Manohar et al., Individual consultant practice does not affect the overall intervention rate: a 6-year study consultant on-call from January 2006 to January 2012. The data from nine out of the 11 consultants were analysed, as two doctors started work part way through the study in 2008. Any protocol changes made between 2008 and the end of the study would be reflected by these two doctor’s practice and hence comparing these data values with the others would make the study unreliable. To account for the possibility of consultants inheriting management decisions made during the previous shift, a secondary database was created where any births that occurred in the first 6 h of a consultant’s shift were removed (the mean duration of labour was 5 h and 38 min). At certain times during the 6-year period, associate specialists were also present in the labour ward and were, therefore, included separately in the analysis. All data were analysed using IBM SPSS Statistics (SPSS Inc., Chicago, IL, USA), to perform the χ2 test and to determine the statistical significance of any observed differences in the data.
Results A total of 20,187 birth records from nine different obstetric consultants were analysed with a secondary analysis of 15,010 births (Table 1). The mean age of mothers in the study was 27.1 years (standard deviation, SD 5.9), the mean BMI was 24.8 (SD 5.5) and the mean weight of infants was 3390 g (SD 5.61). There was no significant difference between consultants in the basic characteristics. Primary analysis (Table 2) showed that there were no statistical differences in the rate of normal vaginal delivery (χ2 = 4.478, P = 0.812) and vacuum delivery (χ2 = 12.232, P = 0.141) between the consultants. However, both the differences in emergency caesarean section rate (χ2 = 24.535, P = 0.02) and forceps rate (χ2 = 21.462, P = 0.006) between the consultants were shown to be statistically significant. It is also interesting to note that when forceps and vacuum numbers were combined under the heading “assisted
deliveries”, there were no significant differences between the consultants. Secondary analysis showed the same interventions (forceps and caesarean) to be significant (Table 3). All other management variables between consultants showed no statistical significance (epidural, episiotomy, neonatal unit (NNU) admission, Apgar score at 1 min and Apgar score at 5 min) in the primary analysis (Table 4). In addition to supporting the findings of the primary analysis, the secondary analysis showed significant differences in the Apgar score at 5 min (P = 0.015, Table 5) amongst the consultants. Assisted deliveries once again showed no significant difference (P = 0.389, Table 5). If the days with associate specialists were compared with the same day without, there was no significant effect on any consultant’s data.
Discussion In some of the fields, data were missing for some patients. The missing data if collected might have resulted in different individual consultant results (such as average BMI). This, however, is unlikely to have changed our overall findings, as there is no reason to suspect that any particular consultant was affected more or less by gaps in the data. The study also assumes that all the data recorded are correct and that there are no errors. Another assumption made in the analysis was negating the impact of the team involved with the consultant in management. The assumption was made that as the teams of one registrar and one junior doctor are equally rotated between the consultants their input into management would affect all individual consultants’ data equally.
Table 1 Patient characteristics mean (95% confidence interval). Consultant
Mean age of mothers (years)
BMI
Gravida
Parity
A B C D E F G H I Average SD P value Number of valid records
27.2 (27.0–27.4) 27.1 (26.9–27.3) 26.9 (26.6–27.1) 26.9 (26.7–27.2) 27.3 (27.0–27.5) 27.1 (26.9–27.4) 27.1 (26.9–27.4) 27.0 (26.7–27.2) 27.1 (26.8–27.3) 27.1 (27.0–27.2) 5.9 0.307 20,121
24.8 (24.6–25.0) 24.6 (24.4–24.9) 24.6 (24.4–24.9) 24.6 (24.4–24.9) 24.8 (24.6–25.1) 24.9 (24.7–25.2) 24.7 (24.5–25.0) 24.8 (24.5–25.1) 24.9 (24.6–25.1) 24.8 (24.7–24.8) 5.5 0.658 16,386
2.3 (2.2–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.2–2.4) 2.3 (2.2–2.3) 2.4 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 1.5 0.545 20,012
0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.8 (0.8–0.9) 0.8 (0.7–0.8) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 1.1 0.230 19,875
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
A B C D E F G H I Average χ P value Number of valid cases
Normal Forceps Vacuum Caesarean Assisted delivery (%) (%) (%) section (%) (%) 83.0 4.4 2.8 81.7 4.8 2.4 83.2 3.8 3.7 84.8 4.2 2.8 82.7 5.1 2.6 83.4 4.6 2.6 83.1 5.3 2.3 83.4 6.7 2.5 83.9 4.8 2.7 83.2 4.8 2.7 6.697 19.799 9.077 0.570 0.011 0.336 15,002 15,002 15,002
9.8 11.1 9.2 8.3 9.6 9.4 9.3 7.4 8.7 9.2 17.912 0.022 15,002
7.2 7.2 7.6 7.0 7.7 7.2 7.6 9.2 7.5 7.6 8.523 0.384 15,002
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
0.966 20,100
0:16 (0:13–0:19) 0:17 (0:13–0:20) 0:14 ( 0:12–0:16) 0:14 (0:12–0:16) 0:15 (0:12–0:18) 0:14 (0:11–0:17) 0:12 (0:10–0:14) 0:12 (0:10–0:13) 0:15 (0:12–0:18) 0:15 (0:13–0:15) 01:00 0.111 20,020 0:46 (0:43–0:48) 0:47 (0:44–0:49) 0:44 (0:42–0:47) 0:47 (0:44–0:49) 0:45 (0:43–0:48) 0:48 (0:45–0:51) 0:46 (0:44–0:49) 0:45 (0:42–0:47) 0:44 (0:41–0:46) 0:46 (0:45–0:47) 01:00 0.446 18,669 4:43 (4:34–4:51) 4:46 (4:37–4:55) 4:45 (4:37–4:54) 4:38 (4:29–4:47) 4:34 (4:25–4:43) 4:36 (4:27–4:46) 4:42 (4:33–4:51) 4:35 (4:26–4:43) 4:41 (4:32–4:50) 4:40 (4:37–4:43) 03:26 0.473 18,657 0.8 1.2 1.1 1.1 0.9 0.6 0.6 0.7 0.5 0.8 12.91 0.115 19,779 5.7 5.6 5.3 5.7 5.2 6.1 5.5 4.7 5.4 5.5 5.086 0.748 19,909 3.9 3.7 3.9 3.3 4.0 3.2 3.7 3.0 3.5 3.6 5.96 0.652 20,080 11.4 10.9 11.4 12.2 11.7 13.0 12.0 12.5 11.3 11.8 7.362 0.498 20,114
Consultant
38.9 37.6 36.0 36.9 37.1 36.4 35.6 36.1 34.9 36.6 10.689 0.220 18,861
Table 3 Secondary analysis: consultants vs. delivery method rates.
Table 4 Primary analysis: other variables between consultants.
Even though the data used were collected over a 6-year period, the amount of data we have may be a limiting factor. We had a cohort of 20,187 women – the caesarean section and assisted deliveries (forceps and vacuum combined) represented approximately 9.4% (1898) and 7.9% (1595) of the study population, respectively. The counter argument for obtaining more data by studying earlier years is that the change in protocol over the years would start impacting the results. This study shows that the rates of normal delivery vs. intervention were not significantly different between consultants managing women with spontaneous onset of labour. However, there was some evidence to suggest that, when an intervention was performed, there was a significant difference in consultant practice in the both
A B C D E F G H I Average SD χ P value Number of valid records
Duration of third stage (h:min)
7.1 7.0 8.1 7.5 8.0 7.9 8.5 9.4 7.9 7.9 12.769 0.120 20,173
Duration of first Duration of second stage (h:min) stage (h:min)
10.4 11.1 9.2 9.1 9.5 9.9 9.4 7.3 8.8 9.4 24.535 0.002 20,173
Episiotomy NNU admissions Apgar score at Apgar score at performed (%) (%) 1 min (%
Sanjay Manohar, Christopher F.G. Woods and Stephen W. Lindow*
Individual consultant practice does not affect the overall intervention rate: a 6-year study Background
Abstract Background: Differences exist in obstetric intervention rates between hospitals but it is not known if the individual consultant governs the decision to intervene or whether intervention is a product of agreed protocols and working practices. The purpose of this study is to analyse the differences in obstetric intervention rates amongst individual consultants working in a large maternity unit. Methods: Each consultant was responsible for all deliveries occurring in successive 24-h periods. Over a 6-year period all deliveries resulting from a spontaneous onset of labour were matched to the consultant in charge at the time of the delivery and analysed. Results: There were no differences seen in normal delivery rates (χ2 = 4.478, P = 0.812) and vacuum (χ2 = 12.232, P = 0.141) rates for the consultants. Significant differences were found in both forceps rate (χ2 = 21.462, P = 0.006) and caesarean rate (χ2 = 24.535, P = 0.002) between consultants. When the forceps rate was combined with vacuum rate there were no significant differences. Conclusions: Within the hospital, individual consultants demonstrated no significant variations in overall intervention rates. However, when intervention occurred, different consultants showed preferences for forceps and caesarean section. Keywords: Caesarean; labour management; specialist obstetrician. DOI 10.1515/jpm-2014-0021 Received January 20, 2014. Accepted April 4, 2014. Previously published online May 8, 2014.
*Corresponding author: Stephen W. Lindow, Senior Lecturer in Perinatology, Women and Children’s Hospital, Hull and East Yorkshire Hospitals, NHS Trust, Kingston upon Hull, East Riding of Yorkshire, UK, Phone: +01482 875875, Fax: +01482 382781, E-mail: [email protected] Sanjay Manohar and Christopher F.G. Woods: Hull York Medical School, Hull Royal Infirmary, Hull, UK
The UK is one of the safest places to give birth, with a maternal mortality rate of 11.39 per 100,000 m aternities [6]. Current obstetric practice in the UK is guided by the Royal College of Obstetricians and Gynaecologists (RCOG) which has outlined standards for clinical practice and issued guidelines for the management of specific conditions [10]. Guidelines are, however, open to interpretation by clinicians who frequently have their own opinions on preferred management options [2]. Labour ward management involves the interaction of senior and junior obstetricians, anaesthetists and midwives on a regular basis [7]. The decisions relating to obstetric intervention could be governed by an individual consultant or by the generally agreed protocols operative in a unit. Previous work has suggested that there are indeed differences in the way patients are treated by health professionals within specialities for similar medical problems [13]. The purpose of this study is to investigate whether obstetric consultants working in the same hospital managed labour differently when compared to one another, in terms of overall intervention rates.
Methods This study has used existing data from the Obstetric department in the Hull and East Yorkshire Women and Children’s hospital. The obstetric department delivers over 5500 live babies every year and consists of 11 consultants who are in turn responsible for 24-h shifts. Consultants are physically present in the labour ward for 8–12 h per day (4 h minimum at weekends) and on call from home the rest of the time. A consultant works with a registrar and senior house officer who are present for 24 h on a shift pattern of work. All caesarean sections must be authorised by a consultant but forceps and vacuum extractions can be authorised and performed by a registrar. The existing data were recorded immediately after each individual delivery by the midwife who was responsible for giving care to the patient. In order to have a fair comparison between consultants, patients who had labour induction of any form, elective caesarean section, twins and stillbirths were removed, leaving only spontaneous onset of labour with single live births for analysis. Each patient’s delivery times were then cross-matched with the corresponding
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
38 Manohar et al., Individual consultant practice does not affect the overall intervention rate: a 6-year study consultant on-call from January 2006 to January 2012. The data from nine out of the 11 consultants were analysed, as two doctors started work part way through the study in 2008. Any protocol changes made between 2008 and the end of the study would be reflected by these two doctor’s practice and hence comparing these data values with the others would make the study unreliable. To account for the possibility of consultants inheriting management decisions made during the previous shift, a secondary database was created where any births that occurred in the first 6 h of a consultant’s shift were removed (the mean duration of labour was 5 h and 38 min). At certain times during the 6-year period, associate specialists were also present in the labour ward and were, therefore, included separately in the analysis. All data were analysed using IBM SPSS Statistics (SPSS Inc., Chicago, IL, USA), to perform the χ2 test and to determine the statistical significance of any observed differences in the data.
Results A total of 20,187 birth records from nine different obstetric consultants were analysed with a secondary analysis of 15,010 births (Table 1). The mean age of mothers in the study was 27.1 years (standard deviation, SD 5.9), the mean BMI was 24.8 (SD 5.5) and the mean weight of infants was 3390 g (SD 5.61). There was no significant difference between consultants in the basic characteristics. Primary analysis (Table 2) showed that there were no statistical differences in the rate of normal vaginal delivery (χ2 = 4.478, P = 0.812) and vacuum delivery (χ2 = 12.232, P = 0.141) between the consultants. However, both the differences in emergency caesarean section rate (χ2 = 24.535, P = 0.02) and forceps rate (χ2 = 21.462, P = 0.006) between the consultants were shown to be statistically significant. It is also interesting to note that when forceps and vacuum numbers were combined under the heading “assisted
deliveries”, there were no significant differences between the consultants. Secondary analysis showed the same interventions (forceps and caesarean) to be significant (Table 3). All other management variables between consultants showed no statistical significance (epidural, episiotomy, neonatal unit (NNU) admission, Apgar score at 1 min and Apgar score at 5 min) in the primary analysis (Table 4). In addition to supporting the findings of the primary analysis, the secondary analysis showed significant differences in the Apgar score at 5 min (P = 0.015, Table 5) amongst the consultants. Assisted deliveries once again showed no significant difference (P = 0.389, Table 5). If the days with associate specialists were compared with the same day without, there was no significant effect on any consultant’s data.
Discussion In some of the fields, data were missing for some patients. The missing data if collected might have resulted in different individual consultant results (such as average BMI). This, however, is unlikely to have changed our overall findings, as there is no reason to suspect that any particular consultant was affected more or less by gaps in the data. The study also assumes that all the data recorded are correct and that there are no errors. Another assumption made in the analysis was negating the impact of the team involved with the consultant in management. The assumption was made that as the teams of one registrar and one junior doctor are equally rotated between the consultants their input into management would affect all individual consultants’ data equally.
Table 1 Patient characteristics mean (95% confidence interval). Consultant
Mean age of mothers (years)
BMI
Gravida
Parity
A B C D E F G H I Average SD P value Number of valid records
27.2 (27.0–27.4) 27.1 (26.9–27.3) 26.9 (26.6–27.1) 26.9 (26.7–27.2) 27.3 (27.0–27.5) 27.1 (26.9–27.4) 27.1 (26.9–27.4) 27.0 (26.7–27.2) 27.1 (26.8–27.3) 27.1 (27.0–27.2) 5.9 0.307 20,121
24.8 (24.6–25.0) 24.6 (24.4–24.9) 24.6 (24.4–24.9) 24.6 (24.4–24.9) 24.8 (24.6–25.1) 24.9 (24.7–25.2) 24.7 (24.5–25.0) 24.8 (24.5–25.1) 24.9 (24.6–25.1) 24.8 (24.7–24.8) 5.5 0.658 16,386
2.3 (2.2–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.2–2.4) 2.3 (2.2–2.3) 2.4 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 2.3 (2.3–2.4) 1.5 0.545 20,012
0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.8 (0.8–0.9) 0.8 (0.7–0.8) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 0.9 (0.8–0.9) 1.1 0.230 19,875
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
A B C D E F G H I Average χ P value Number of valid cases
Normal Forceps Vacuum Caesarean Assisted delivery (%) (%) (%) section (%) (%) 83.0 4.4 2.8 81.7 4.8 2.4 83.2 3.8 3.7 84.8 4.2 2.8 82.7 5.1 2.6 83.4 4.6 2.6 83.1 5.3 2.3 83.4 6.7 2.5 83.9 4.8 2.7 83.2 4.8 2.7 6.697 19.799 9.077 0.570 0.011 0.336 15,002 15,002 15,002
9.8 11.1 9.2 8.3 9.6 9.4 9.3 7.4 8.7 9.2 17.912 0.022 15,002
7.2 7.2 7.6 7.0 7.7 7.2 7.6 9.2 7.5 7.6 8.523 0.384 15,002
Brought to you by | Karolinska Institute Authenticated Download Date | 5/30/15 12:31 PM
0.966 20,100
0:16 (0:13–0:19) 0:17 (0:13–0:20) 0:14 ( 0:12–0:16) 0:14 (0:12–0:16) 0:15 (0:12–0:18) 0:14 (0:11–0:17) 0:12 (0:10–0:14) 0:12 (0:10–0:13) 0:15 (0:12–0:18) 0:15 (0:13–0:15) 01:00 0.111 20,020 0:46 (0:43–0:48) 0:47 (0:44–0:49) 0:44 (0:42–0:47) 0:47 (0:44–0:49) 0:45 (0:43–0:48) 0:48 (0:45–0:51) 0:46 (0:44–0:49) 0:45 (0:42–0:47) 0:44 (0:41–0:46) 0:46 (0:45–0:47) 01:00 0.446 18,669 4:43 (4:34–4:51) 4:46 (4:37–4:55) 4:45 (4:37–4:54) 4:38 (4:29–4:47) 4:34 (4:25–4:43) 4:36 (4:27–4:46) 4:42 (4:33–4:51) 4:35 (4:26–4:43) 4:41 (4:32–4:50) 4:40 (4:37–4:43) 03:26 0.473 18,657 0.8 1.2 1.1 1.1 0.9 0.6 0.6 0.7 0.5 0.8 12.91 0.115 19,779 5.7 5.6 5.3 5.7 5.2 6.1 5.5 4.7 5.4 5.5 5.086 0.748 19,909 3.9 3.7 3.9 3.3 4.0 3.2 3.7 3.0 3.5 3.6 5.96 0.652 20,080 11.4 10.9 11.4 12.2 11.7 13.0 12.0 12.5 11.3 11.8 7.362 0.498 20,114
Consultant
38.9 37.6 36.0 36.9 37.1 36.4 35.6 36.1 34.9 36.6 10.689 0.220 18,861
Table 3 Secondary analysis: consultants vs. delivery method rates.
Table 4 Primary analysis: other variables between consultants.
Even though the data used were collected over a 6-year period, the amount of data we have may be a limiting factor. We had a cohort of 20,187 women – the caesarean section and assisted deliveries (forceps and vacuum combined) represented approximately 9.4% (1898) and 7.9% (1595) of the study population, respectively. The counter argument for obtaining more data by studying earlier years is that the change in protocol over the years would start impacting the results. This study shows that the rates of normal delivery vs. intervention were not significantly different between consultants managing women with spontaneous onset of labour. However, there was some evidence to suggest that, when an intervention was performed, there was a significant difference in consultant practice in the both
A B C D E F G H I Average SD χ P value Number of valid records
Duration of third stage (h:min)
7.1 7.0 8.1 7.5 8.0 7.9 8.5 9.4 7.9 7.9 12.769 0.120 20,173
Duration of first Duration of second stage (h:min) stage (h:min)
10.4 11.1 9.2 9.1 9.5 9.9 9.4 7.3 8.8 9.4 24.535 0.002 20,173
Episiotomy NNU admissions Apgar score at Apgar score at performed (%) (%) 1 min (%
