Objective
We sought to examine the difference in use of labor epidural analgesia among women from different neighborhood socioeconomic groups.
Study Design
Neighborhood socioeconomic variables from the 2001 Canadian Census were linked to singleton vaginal births from the Niday perinatal database (2004–2006) in Ontario, Canada. Births were divided into income and education groups by quintiles. Generalized estimating equations were employed to evaluate the association between labor epidural and neighborhood socioeconomic status. Supplementary analysis was conducted after stratifying data by hospital types.
Results
Compared with those from the richest neighborhood, women from the poorest quintile were the least likely to receive labor epidural analgesia (odds ratio [OR], 0.59; 95% confidence interval [CI], 0.58–0.61). The differences were smallest in teaching hospitals (OR, 0.73; 95% CI, 0.67–0.79) and largest in small community hospitals (OR, 0.57; 95% CI, 0.50–0.64). Similar association was found in neighborhood education quintiles.
Conclusion
The use of labor epidural analgesia is decreased with decreasing neighborhood economic and education levels.
Epidural analgesia has been proved to be more effective than most other methods of pain relief during labor. The rate of intrapartum epidural analgesia increased dramatically in recent years in many industrialized countries. It continuously increased from 22% in 1981 to 61% in 2001 in the United States. In Sweden, approximately 42% of primiparous women used epidural anesthesia in 2000, nearly doubled that of the figure in 1995 (22%). In Ontario, Canada, the rate also increased from 34–38% during an 11-year period of 1982–1993.
Although the increase in use of epidural analgesia in managing labor pain has been regarded as one of the most significant achievements of modern obstetric practice, not every woman has access to it. Previous studies have shown that the use of epidural analgesia varied significantly among people of different socioeconomic status (SES). For example, studies in the United States indicated that the use of epidural during labor was associated with not only clinical characteristics, such as parity, plurality, and clinical risk factors, but also women’s ethnicity, insurance status, and rural-urban residence. In Australia, women from rural areas had lower rates of obstetric interventions, including intrapartum epidural use, than those from urban areas. In Germany, intrapartum epidural analgesia was reported to be affected by insurance type, maternal occupation and education degree, and immigration status. All these studies concluded that socioeconomic disadvantage was associated with lower rate of epidural use in labor.
The Canadian health care system is a universal one, and the government provides essential health services to all people. Previous studies have reported that SES did not affect Canadian women’s access to obstetric care services, such as labor induction and cesarean delivery. However, a study from Israel, where basic health care including epidural analgesia and other procedures in childbirth were fully covered by national health insurance, revealed that highly educated and secular (nonreligious belief) women were more likely to accept intrapartum epidural analgesia. To our knowledge, no study in Canada has examined whether the use of epidural in labor differs among women from different socioeconomic groups.
The main purpose of the present study is to examine whether the use of epidural analgesia in labor differs among women from different neighborhood socioeconomic strata in a publicly funded health care system. Moreover, the article attempts to find out whether this association is consistent among women delivering babies in different types of hospitals.
Materials and Methods
For this study, we used the 2004–2006 Niday perinatal database of the province of Ontario, Canada. This Internet-based regional perinatal surveillance database was hosted by the CritiCall Ontario. By 2006, a total of 82 hospitals and midwifery practice groups reported to the database, enabling it to cover >95% of births in the province of Ontario. The Niday database provided information on maternal and prenatal characteristics, health service status, intrapartum interventions, and infant outcomes.
We linked the Niday perinatal database to the 2001 Canadian Census data for Ontario by converting maternal postal codes into dissemination areas (DAs), which were small, relatively stable geographic units with a population of 400–700 persons of relatively homogenous SES. We excluded records with missing information on birth outcomes, pain relief, multiple gestations, nonhospital deliveries, and cesarean section deliveries.
Outcome of interest in this study was epidural analgesia as a method of pain relief for this delivery.
Birth records were first sorted by the family income in their corresponding DAs and classified into 5 levels according to neighborhood income quintiles. Birth records were then arranged likewise based on the percentage of people acquiring at least college education in each DA and categorized into neighborhood education quintiles.
Individual-level covariances used for analysis included hospital of delivery (small community hospitals, large community hospitals, and teaching hospitals); maternal age at delivery; parity; smoking during pregnancy; having obstetricians as antenatal health providers, and preexisting maternal health problems (chronic hypertension, diabetes, heart disease, thyroid disease, lupus, alcoholism, asthma, human immunodeficiency virus, mental illness, hepatitis B, drug use, psychiatric disorders, depression, anxiety, and other diseases) present or not. Individual-level characteristics of this delivery included gestation age, infant birthweight, obstetric complications of this pregnancy, intrapartum complications during labor, instrumental delivery, and labor type.
We first compared the distributions of maternal and infant characteristics among neighborhood income and education quintiles by using frequency as descriptive statistics and Pearson χ 2 test in comparison. The prevalence of epidural use in different neighborhood socioeconomic strata and according to maternal and infant characteristics was reported both in total and stratified by hospital type, with frequency as descriptive statistics and χ 2 test in comparison. The association between neighborhood socioeconomic strata and epidural use was estimated by using generalized estimating equations (GEEs) to adjust for potential confounders at individual level. Individual-level variables considered in GEE models included maternal age at delivery (<20, 20–34, and ≥35 years), parity (1, 2, and ≥3), smoking during pregnancy (yes/no), having obstetricians as antenatal health providers (yes/no), preexisting maternal health problems (present or absent), gestation (<37 or ≥37 weeks), infant birthweight (<2500, 2500–4000, and ≥4000 g), obstetric complications of this pregnancy (present or absent), intrapartum complications during labor (present or absent), instrumental delivery (yes/no), and labor type (induced or spontaneous). To assess the effect modification by hospital of delivery, the association between neighborhood SES and labor epidural was further analyzed after stratifying all data by hospital type.
To reduce loss of samples in regression analysis, missing values >10% in independent variables were coded as a separate category in GEE models. Sensitivity analysis using records without incomplete information was conducted to assess the possible impacts of including those records with missing information on some confounding factors. The analysis was performed by using Statistical Analysis System, version 9.1 (SAS Institute Inc, NC). An alpha of .05 was used to determine statistical significance.
Results
A total of 220,814 vaginal singleton births were included in analysis. There was high correlation ( P < .0001) but low consistency (Kappa = 0.32; 95% confidence interval [CI], 0.31–0.33) between neighborhood income and education quintiles. With the increase of neighborhood income level, the proportion of women who gave birth in small community hospitals decreased, while the proportion of women who gave birth in teaching hospitals increased. Similar trend was also observed in neighborhood education quintiles. Both lower neighborhood income and lower neighborhood education were associated with higher percentage of teenaged mothers, lower percentage of advanced maternal age, higher gravidity, or tobacco use during pregnancy; preexisting health problems; preterm birth (<37 weeks); low infant birthweight (<2500 g); or intrapartum complications. With the increase of neighborhood education level, the proportion of women having obstetricians as antenatal health provider increased, while the proportion of mothers having obstetric complications and instrumental delivery decreased ( Table 1 ).
| Characteristics | Total | Neighborhood family income quintiles | Neighborhood education quintiles | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 (Poorest) | 2 | 3 | 4 | 5 (Richest) | 1 (Lowest) | 2 | 3 | 4 | 5 (Highest) | ||
| No. of subjects | 220,814 | 43,863 | 44,795 | 44,560 | 44,150 | 43,446 | 45,317 | 45,109 | 43,689 | 44,038 | 42,661 |
| Hospital of delivery | |||||||||||
| Small community | 9.70 | 9.58 | 14.36 | 12.69 | 7.06 | 4.65 | 20.05 | 12.97 | 8.03 | 5.20 | 1.62 |
| Large community | 76.49 | 76.30 | 73.98 | 74.70 | 80.91 | 76.62 | 69.63 | 77.50 | 80.74 | 81.44 | 73.26 |
| Teaching | 13.81 | 14.11 | 11.66 | 12.61 | 12.03 | 18.73 | 10.32 | 9.53 | 11.23 | 13.36 | 25.12 |
| Maternal age group, y | |||||||||||
| <20 | 3.98 | 6.99 | 5.19 | 3.69 | 2.45 | 1.56 | 7.52 | 4.97 | 3.52 | 2.42 | 1.27 |
| 20–34 | 77.93 | 78.72 | 80.47 | 79.73 | 78.50 | 72.09 | 80.27 | 80.25 | 79.51 | 77.43 | 71.89 |
| ≥35 | 18.09 | 14.29 | 14.33 | 16.58 | 19.05 | 26.36 | 12.21 | 14.78 | 16.97 | 20.16 | 26.84 |
| Parity | |||||||||||
| 1 | 45.61 | 45.44 | 47.24 | 46.05 | 45.65 | 43.59 | 43.06 | 44.14 | 45.37 | 46.80 | 48.90 |
| 2 | 34.43 | 31.37 | 32.81 | 34.61 | 35.70 | 37.72 | 32.60 | 34.59 | 35.01 | 35.19 | 34.81 |
| ≥3 | 19.97 | 23.19 | 19.96 | 19.34 | 18.65 | 18.70 | 24.34 | 21.27 | 19.62 | 18.01 | 16.29 |
| Smoking during pregnancy | n = 195,142 | 38,127 | 39,545 | 39,564 | 39,145 | 38,761 | 41,221 | 39,520 | 38,111 | 38,949 | 37,341 |
| No | 88.20 | 82.12 | 84.55 | 87.97 | 91.79 | 94.52 | 78.80 | 85.33 | 89.49 | 92.64 | 95.67 |
| Yes | 11.80 | 17.88 | 15.45 | 12.03 | 8.21 | 5.48 | 21.20 | 14.67 | 10.51 | 7.36 | 4.33 |
| Maternal health problems a | n = 159,888 | 32,465 | 33,431 | 32,352 | 31,370 | 30,270 | 35,202 | 33,070 | 30,903 | 30,796 | 29,917 |
| No | 80.40 | 79.60 | 78.91 | 79.90 | 81.83 | 81.96 | 76.49 | 79.41 | 81.40 | 81.79 | 83.76 |
| Yes | 19.60 | 20.40 | 21.09 | 20.10 | 18.17 | 18.04 | 23.51 | 20.59 | 18.60 | 18.21 | 16.37 |
| Obstetrician as antenatal health provider | n = 162,142 | 32,198 | 33,880 | 32,980 | 32,046 | 31,038 | 35,247 | 33,424 | 31,581 | 31,525 | 30,365 |
| No | 23.03 | 21.08 | 25.42 | 25.43 | 22.19 | 20.77 | 26.21 | 24.11 | 22.18 | 21.39 | 20.74 |
| Yes | 76.97 | 78.92 | 74.58 | 74.57 | 77.81 | 79.23 | 73.79 | 75.59 | 77.82 | 78.61 | 79.26 |
| Gestation, wk | |||||||||||
| <37 | 6.02 | 6.55 | 6.22 | 6.12 | 5.78 | 5.43 | 6.54 | 6.38 | 5.92 | 5.89 | 5.32 |
| ≥37 | 93.98 | 93.46 | 93.78 | 93.88 | 94.22 | 94.57 | 93.46 | 93.62 | 94.08 | 94.11 | 94.68 |
| Infant birthweight, g | |||||||||||
| <2500 | 4.56 | 5.65 | 4.76 | 4.55 | 4.05 | 3.77 | 5.09 | 4.80 | 4.56 | 4.42 | 3.88 |
| 2500–4000 | 84.83 | 85.37 | 84.83 | 84.36 | 84.84 | 84.77 | 83.61 | 84.50 | 85.00 | 85.19 | 85.93 |
| >4000 | 10.61 | 8.98 | 10.41 | 11.10 | 11.11 | 11.45 | 11.30 | 10.70 | 10.44 | 10.39 | 10.18 |
| Obstetric complications b | n = 160,121 | 32,308 | 33,630 | 32,531 | 31,471 | 30,181 | 35,386 | 33,167 | 30,957 | 30,749 | 29,862 |
| No | 79.74 | 78.91 | 78.74 | 79.38 | 80.68 | 81.14 | 77.43 | 78.94 | 80.19 | 80.01 | 82.63 |
| Yes | 20.26 | 21.09 | 21.26 | 20.62 | 19.32 | 18.86 | 22.57 | 21.06 | 19.81 | 19.99 | 17.37 |
| Intrapartum complications c | n = 164,874 | 33,022 | 34,396 | 33,503 | 32,623 | 31,330 | 36,339 | 34,091 | 31,918 | 31,812 | 30,714 |
| No | 74.25 | 73.03 | 74.08 | 74.60 | 75.21 | 74.35 | 75.29 | 74.99 | 74.52 | 73.34 | 72.88 |
| Yes | 25.75 | 26.97 | 25.92 | 25.40 | 24.79 | 25.65 | 24.71 | 25.01 | 25.48 | 26.66 | 27.12 |
| Instrumental delivery d | |||||||||||
| No | 85.03 | 84.95 | 85.61 | 85.42 | 84.65 | 84.47 | 87.43 | 86.11 | 84.87 | 84.16 | 82.28 |
| Yes | 14.97 | 15.05 | 14.39 | 14.58 | 15.35 | 15.53 | 12.57 | 13.89 | 15.13 | 15.84 | 17.72 |
| Labor type | |||||||||||
| Induced | 26.76 | 25.70 | 27.17 | 27.33 | 26.49 | 27.10 | 29.03 | 27.49 | 26.07 | 25.49 | 25.60 |
| Spontaneous | 73.24 | 74.30 | 72.83 | 72.67 | 73.51 | 72.90 | 70.97 | 72.51 | 73.93 | 74.51 | 74.40 |
| Obstetrician-assisted delivery | |||||||||||
| No | 18.21 | 16.25 | 19.92 | 20.10 | 17.64 | 17.04 | 21.12 | 19.09 | 17.29 | 16.79 | 16.58 |
| Yes | 81.79 | 83.75 | 80.08 | 79.90 | 82.36 | 82.96 | 78.88 | 80.91 | 82.71 | 83.21 | 83.42 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
