Continuing Epidural Post Cesarian Opioid Dependent

• Journal List
• HHS Author Manuscripts
• PMC8599660

Am J Obstet Gynecol MFM. Author manuscript; available in PMC 2021 Nov 18.

Published in final edited form as:

PMCID: PMC8599660

NIHMSID: NIHMS1747019

The risk of serious opioid-related events associated with common opioid prescribing regimens in the postpartum period after cesarean delivery

Andrew D. Wiese, PhD, MPH, Sarah S. Osmundson, MD, MS, Edward Mitchel, Jr, MS, Margaret Adgent, PhD, MSPH, Sharon Phillips, MSPH, Stephen W. Patrick, MD, MPH, MS, Andrew J. Spieker, PhD, and Carlos G. Grijalva, MD, MPH

Andrew D. Wiese

Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN

Sarah S. Osmundson

Obstetrics and Gynecology, Vanderbilt University Medical Center, Nashville, TN

Edward Mitchel, Jr

Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN

Margaret Adgent

Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN

Sharon Phillips

Biostatistics, Vanderbilt University Medical Center, Nashville, TN

Stephen W. Patrick

Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN; The Center for Child Health Policy, Vanderbilt University Medical Center, Nashville, TN

Andrew J. Spieker

Biostatistics, Vanderbilt University Medical Center, Nashville, TN

Carlos G. Grijalva

Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN ; Geriatric Research, Education and Clinical Center, Tennessee Valley Healthcare System, US Department of Veterans Affairs, Nashville, TN

Abstract

BACKGROUND:

Opioid analgesics are commonly prescribed to women after cesarean delivery. There is a growing effort to prescribe opioids judiciously; however, the risk of serious opioid-related events associated with specific prescribing patterns after cesarean delivery remains unclear.

OBJECTIVE:

We examined the association between the dosage of the first opioid prescription filled after cesarean delivery and the risk of serious opioid-related events.

STUDY DESIGN:

We identified opioid-naïve women with a cesarean delivery enrolled in Tennessee Medicaid (2007–2014). Pharmacy prescription fill data characterized opioids filled within 5 days after delivery. Patients were followed up from day 5 after delivery to the earliest of the following: serious opioid-related event (persistent opioid use, evidence of opioid use disorder [diagnosis or methadone or buprenorphine fill], overdose, or opioid-related death), non–opioid-related death, enrollment loss, or 365th day. We estimated the adjusted hazard ratios and 95% confidence intervals for the serious opioid-related event outcomes based on the dosage (morphine milligram equivalents) of the first filled opioid prescription, adjusting for baseline sociodemographic characteristics, delivery complications, multiple deliveries, comorbidities, and medication use. Secondary analyses examined the role of commonly prescribed opioid strengths and quantities.

RESULTS:

The overall incidence rate of serious opioid-related events among women after cesarean delivery was 3.0 per 100 person-years. Compared with women who did not fill an opioid prescription, the rate of serious opioid-related events was higher among women who filled an opioid prescription, although only significantly higher among women who filled a total dosage of ≥100 morphine milligram equivalents (97.1% of opioid prescriptions). In the secondary analyses, women with a low prescribed daily opioid dosage and women with a low prescribed number of oxycodone (5 mg) tablets (<10 tablets) were not at increased risk of serious opioid-related events compared with women who did not fill an opioid prescription.

CONCLUSION:

Opioid-naïve women who filled a postpartum opioid prescription at commonly prescribed doses after cesarean delivery had an increased risk of serious opioid-related events compared to women who did not fill a postpartum opioid prescription. Low opioid doses were not associated with a significant increase in the risk of serious opioid-related events.

Keywords: cesarean delivery, dose, opioid analgesics, opioid-related death, opioid-related overdose, persistent opioid use, postpartum, substance use disorder

Introduction

Cesarean delivery remains the most common major surgery in the United States with >1.2 million cesarean deliveries performed annually.¹ More than 75% of women fill an opioid prescription after cesarean delivery.^2,3 For many women, postpartum opioid prescribing is their first exposure to opioids. Such exposure is a potential precipitating event for the development of serious opioid-related events (SOREs), including persistent opioid use, opioid use disorder (OUD), and opioid-related overdose and death.^4-9 Increasing evidence indicates that for both vaginal and cesarean deliveries, opioid prescribing during the postpartum period is associated with an increased risk of SOREs.^2,5,6,10

Approaches that improve postpartum opioid prescribing, including after cesarean delivery, are greatly needed.^11,12 Currently, opioid prescribing guidelines for acute pain among opioid-naïve patients recommend the use of the lowest effective opioid dose for the shortest duration.¹³ However, the risk of SOREs associated with commonly prescribed opioid prescriptions after cesarean delivery remains unclear. Here, we investigated whether the risk of SOREs varied by the dose of the first postpartum opioid prescription received after cesarean delivery.

Materials and Methods

Data sources

We identified a retrospective cohort of women enrolled in Tennessee Medicaid (TennCare) aged 15 to 44 years who underwent a cesarean delivery from 2007 to 2014 in a Tennessee hospital. TennCare, the managed Medicaid program in Tennessee, provides insurance coverage for over half of all annual Tennessee births. The TennCare data include enrollment, claims, and pharmacy files that are used to identify patient eligibility, healthcare encounters, and health conditions and records of filled prescriptions. We supplemented the TennCare data with linked Tennessee vital records data (birth and death certificates) and the Tennessee Hospital Discharge Data System (a registry of all hospital-based encounters in Tennessee).^6,11,14 The study was approved by the institutional review boards of Vanderbilt University and the Tennessee Department of Health and the Division of TennCare.

Study cohort and follow-up

Women with a cesarean delivery linked to a birth certificate (2007–2014) were included in the retrospective cohort. Women were required to be continuously enrolled in TennCare during a baseline period (180 days before delivery through day 4 after delivery [inclusive]), to have been discharged on or before day 4 after delivery, and to be opioid naïve (≤1 before outpatient opioid prescription) during the baseline period. We excluded women with baseline evidence of an OUD (International Classification of Diseases, Ninth Revision [ICD-9], coded diagnosis [304.x, 304.7x, 305.5x, and 965.0] or methadone or buprenorphine fill) (Figure 1).

Study population flow diagram

OUD, opioid use disorder.

Wiese. Postpartum opioid use after cesarean delivery and the risk of opioid-related harms. Am J Obstet Gynecol MFM 2021.

Women meeting cohort eligibility criteria began follow-up on day 5 after delivery (t₀) to allow an adequate period for the identification of a postpartum opioid prescription fill after delivery and discharge (Supplemental Figure 1). Follow-up continued from t₀ to the earliest of the following: day 365 after delivery, loss of enrollment (defined as reaching day 30 without continuous enrollment), non-opioid–related death, or a study outcome (provided in the subsection "Serious opioid-related events").

Exposure

The exposure of interest was the dose of the first outpatient opioid prescription filled in the postpartum period (this encompassed the time from day 3 before delivery to day 4 after delivery). Prescriptions filled in the 3 days before delivery were included to account for women filling an outpatient opioid prescription in preparation for pain management after delivery. Study opioid analgesics included only oral formulations intended for pain management (ie, excluded cough suppressant and antidiarrheal formulations). The total dose of the first postpartum opioid prescription was calculated as the total morphine milligram equivalents (MME) filled (ie, MME=[strength per tablet/solution] × [quantity of tablet/solution] × [MME conversion factor]).¹⁵

In the primary analysis, we included a binary indicator variable for filling a postpartum opioid or not and used restricted cubic splines to model the total opioid dose as a continuous term among those who filled an opioid prescription.^16,17 Modeling a continuous exposure variable without restricted cubic splines would require the restrictive assumption that the dose-response curve is linear. We used restricted cubic splines to model opioid dose as a continuous variable while accommodating a possible nonlinear relationship with the risk of SOREs¹⁸

Covariates

We measured relevant covariates identified a priori during the baseline period, including demographics, clinical conditions potentially associated with increased pain and opioid use, contraindications to the use of nonsteroid antiinflammatory drugs, severe maternal morbidity, bilateral tubal ligation, perineal laceration, and mental health conditions. Clinical and mental health conditions were identified using filled prescriptions for relevant drugs and coded diagnoses. Coded diagnoses required a diagnostic code present during a single hospital visit, 23-hour hospital stay, or emergency department visit or from 2 outpatient encounters on separate dates.

Serious opioid-related events

SOREs were examined as a composite outcome identified on the earliest date one of the following events occurred during follow-up: development of persistent opioid use (filling a 90-day opioid supply within a 180-day interval without a gap in supply of >30 days), development of an OUD (prescription fill for methadone or buprenorphine or coded diagnosis [ICD-9: 304.0x, 304.7x, 305.5x, and 965.0]), opioid-related overdose (ICD-9: 965.0x or E850.0–850.2), or opioid-related death (underlying cause recorded on the death certificate [International Classification of Diseases, Tenth Revision (ICD-10): T40.0–T40.4, X42, X62, and Y12]).^5,6

Statistical analysis

We used Cox proportional hazards regression to compare the risk of SOREs among women who filled a postpartum opioid prescription with that of women that did not, accounting for covariates. We calculated adjusted hazard ratios (aHR) with 95% confidence intervals (CI) using robust standard errors to account for women contributing >1 cesarean delivery to the cohort. We modeled the dose using restricted cubic spline terms using an initial knot at the 10th percentile (125 MME) to assess low dosage use and subsequent interior quintiles of the dose distribution (20%: 150; 40% and 60%: 225; 80%: 270; 99%: 600). Using this model, which included a binary indicator variable for filling a postpartum opioid and all prespecified covariates, we calculated the predicted aHRs at each observed value of the dose up to the 99th percentile of the dose distribution (600 MME), using women that did not fill a postpartum opioid prescription as the reference. We applied sequential multivariate imputation (50 imputed data sets) to account for missing covariate information for a small proportion of birth certificate variables (3.8% of patients).¹⁹ We assessed the proportional hazards assumption by including an interaction between the exposure and log (time).

A secondary analysis examined the daily MME dose as the exposure of interest. We modeled an initial knot at the 10th percentile (30 MME per day) to assess low dosage use and subsequent quartiles of the dosage distribution (37.5, 50, 75, and 112.5 MME per day). To provide more pragmatic information for clinical prescribing of specific opioid types, we conducted a post hoc analysis to assess the risk of SOREs associated with different strength formulations (≤5 mg or >5 mg) and opioid types (hydrocodone and oxycodone; both accounted for 95.8% of all prescriptions). For each opioid type and strength formulation, we used restricted cubic spline terms to model the tablet quantity and days' supply and calculated the predicted aHRs at selected tablet and days' supply values, with women that did not fill a postpartum opioid prescription as reference.

To enhance the identification of women with previous opioid use history, we conducted a sensitivity analysis that included only women with a full year of baseline data available (ie, met all eligibility criteria in the 365 days before delivery through day 4 after delivery). In another sensitivity analysis, we used an alternate composite outcome definition that included only opioid use disorder, opioid-related overdose, and death. A separate sensitivity analysis determined the robustness of our findings to using a postpartum opioid fill window defined based on the discharge date (day 3 before delivery to day 2 after discharge) rather than the cesarean delivery date to address the potential bias from being unable to detect the initial postpartum opioid prescription among women discharged on day 4 after delivery. Furthermore, we conducted a dose-response assessment by repeating the primary analysis restricted to mothers that filled a postpartum opioid prescription. Finally, we examined the subhazard ratios for opioid doses, categorized based on quartiles, using the Fine and Gray method to examine the possibility that non–opioid-related death was a competing risk.²⁰ Statistical analyses were completed using Stata statistical software (release 16.1 and 17.0; Stata-Corp LLC, College Station, TX).

Results

Study population

Among 65,170 women with cesarean delivery of a live infant (Figure 1), the average age at delivery was 24.7 years (median, 24), and most deliveries were singleton births (97.1%) among White women (61.4%). The geographic distribution of cesarean delivery was similar across the regions of Tennessee (West, 31.4%; Central, 36.4%; and East, 32.0%) and by parity (0 previous delivery, 38.6%; 1 previous delivery, 32.6%; and ≥2 previous deliveries, 28.5%). A small percentage of women were missing covariates from linked birth certificates (n=2463 [3.8%]) (Table 1).

TABLE 1

Baseline and follow-up characteristics of women with cesarean delivery stratified by overall opioid morphine milligram equivalent quartiles of the first postpartum opioid prescription, Tennessee Medicaid (2007–2014)

Characteristica	No opioid n=8447		1−<150 MME n=6948		150−<225 MME n=15,572		225 MME n=21,953		>225 MME n=12,250
	n	%	n	%	n	%	n	%	n	%
Year of delivery
2007	1124	13.3	1169	16.8	2066	13.3	2545	11.6	1220	10.0
2008	959	11.4	1052	15.1	2063	13.2	2841	12.9	1510	12.3
2009	911	10.8	1098	15.8	2047	13.1	2815	12.8	1491	12.2
2010	872	10.3	967	13.9	1952	12.5	2883	13.1	1855	15.1
2011	1022	12.1	835	12.0	2010	12.9	2777	12.6	1666	13.6
2012	1079	12.8	807	11.6	2055	13.2	2965	13.5	1660	13.6
2013	1421	16.8	602	8.7	1955	12.6	2978	13.6	1553	12.7
2014	1059	12.5	418	6.0	1424	9.1	2149	9.8	1295	10.6
Race and ethnicity
Non-Hispanic White	4977	58.9	3872	55.7	9150	58.8	13,920	63.4	8121	66.3
Black, Hispanic, or other	3467	41.0	3073	44.2	6416	41.2	8015	36.5	4126	33.7
Parity
0	3099	36.7	2774	39.9	6153	39.5	8467	38.6	4660	38.0
1	2873	34.0	2149	30.9	5048	32.4	7116	32.4	4067	33.2
≥2	2434	28.8	2006	28.9	4344	27.9	6307	28.7	3477	28.4
Region
West	2656	31.4	2723	39.2	5877	37.7	5414	24.7	3759	30.7
Central	2947	34.9	2483	35.7	4966	31.9	8092	36.9	5211	42.5
East	2745	32.5	1732	24.9	4692	30.1	8410	38.3	3261	26.6
Plurality
Singleton	8161	96.6	6780	97.6	15,134	97.2	21,319	97.1	11,875	96.9
Multiple	286	3.4	168	2.4	437	2.8	634	2.9	375	3.1
Income (categorical)
$0–$34,999	1576	18.7	1454	20.9	3352	21.5	4587	20.9	2145	17.5
$35,000–$44,999	2023	23.9	1753	25.2	3952	25.4	5468	24.9	3051	24.9
$45,000–$59,999	1986	23.5	1670	24.0	3742	24.0	5165	23.5	2859	23.3
≥$60,000	2483	29.4	1850	26.6	4016	25.8	6003	27.3	3792	31.0
Delivery complications
Severe maternal morbidityb	271	3.2	161	2.3	388	2.5	623	2.8	337	2.8
Bilateral tubal ligation	1659	19.6	1307	18.8	3083	19.8	4619	21.0	2541	20.7
Laceration	2	0.0	3	0.0	1	0.0	1	0.0	0	0.0
Clinical conditions and medication use
Tobacco use	1774	21.0	1775	25.5	4146	26.6	6189	28.2	3614	29.5
Predelivery opioid use	850	10.1	1045	15	2284	14.7	3120	14.2	1971	16.1
Benzodiazepine use	146	1.7	171	2.5	412	2.6	642	2.9	458	3.7
Stimulant use	22	0.3	20	0.3	49	0.3	59	0.3	44	0.4
Depression medication use	329	3.9	320	4.6	869	5.6	1449	6.6	830	6.8
Psychosis medication use	68	0.8	36	0.5	86	0.6	160	0.7	102	0.8
Depression disorders	232	2.7	102	1.5	309	2.0	731	3.3	329	2.7
Anxiety disorders	128	1.5	64	0.9	197	1.3	370	1.7	195	1.6
Psychosis disorders	32	0.4	5	0.1	19	0.1	39	0.2	13	0.1
Bipolar disorder	167	2.0	62	0.9	193	1.2	408	1.9	192	1.6
NSAID contraindication	27	0.3	7	0.1	34	0.2	53	0.2	37	0.3
Abdominal pain	199	2.4	123	1.8	361	2.3	462	2.1	224	1.8
Dental pain	14	0.2	5	0.1	23	0.1	51	0.2	18	0.1
Genitourinary pain	408	4.8	204	2.9	684	4.4	1291	5.9	501	4.1
Trauma diagnosis	224	2.7	206	3.0	476	3.1	688	3.1	441	3.6
Musculoskeletal pain	365	4.3	336	4.8	701	4.5	862	3.9	590	4.8
Back pain	478	5.7	391	5.6	963	6.2	1400	6.4	736	6.0
Autoimmunity disorder	36	0.4	15	0.2	30	0.2	61	0.3	23	0.2
Arthritis	439	5.2	370	5.3	892	5.7	1282	5.8	702	5.7
Neck pain	17	0.2	9	0.1	23	0.1	29	0.1	16	0.1
Neurologic pain	38	0.4	25	0.4	53	0.3	96	0.4	44	0.4
Malignancy	117	1.4	123	1.8	204	1.3	304	1.4	169	1.4
Irritable bowel syndrome	63	0.7	30	0.4	96	0.6	142	0.6	73	0.6
Headache	286	3.4	196	2.8	506	3.2	869	4.0	434	3.5
Migraine	72	0.9	36	0.5	90	0.6	204	0.9	94	0.8
Pain, not otherwise specified	113	1.3	57	0.8	152	1.0	243	1.1	160	1.3
Other pain disorders	147	1.7	88	1.3	217	1.4	379	1.7	207	1.7
Hospitalization days after the date of delivery
0–1	169	2.0	128	1.8	245	1.6	310	1.4	274	2.2
2	3720	44.0	3832	55.2	8229	52.8	10,886	49.6	6245	51.0
3	3425	40.5	2679	38.6	6345	40.7	9281	42.3	4964	40.5
4	1133	13.4	309	4.4	753	4.8	1476	6.7	767	6.3
Hospitalizations during pregnancyc
≤1	5517	65.3	4195	60.4	9647	62	13,437	61.2	7261	59.3
2	1644	19.5	1515	21.8	3235	20.8	4582	20.9	2581	21.1
≥3	1286	15.2	1238	17.8	2690	17.3	3934	17.9	2408	19.7
Emergency department visits during pregnancyc
0	4979	58.9	4031	58.0	8786	56.4	11,819	53.8	6640	54.2
1	1897	22.5	1716	24.7	3849	24.7	5509	25.1	3073	25.1
≥2	1571	18.6	1201	17.3	2937	18.9	4625	21.1	2537	20.7
Outpatient visits during pregnancyc,d
0	2929	34.7	1665	24.0	3759	24.1	5647	25.7	3131	25.6
1–3	3471	41.1	3399	48.9	7384	47.4	10,438	47.5	5764	47.1
≥4	2047	24.2	1884	27.1	4429	28.4	5868	26.7	3355	27.4
Demographics
Distance to hospital (miles)
Median (IQR)	8.7	(4.1–17.9)	7.9	(4.1–14.7)	8.2	(4.1–15.7)	8.4	(3.8–16.7)	8.8	(3.8–17.7)
Follow-up
Total person-years	7099		5767		12,945		18,216		10,040
Mean follow-up (mo)	10.1		10.0		10.0		10.0		9.8
Age (y)
Mean±standard deviation	25.4	±5.6	24.4	±5.2	24.5	±5.3	24.6	±5.4	24.7	±5.3
Median (IQR)	24.0	(21.0–29.0)	23.0	(20.0–27.0)	23.0	(20.0–28.0)	23.0	(21.0–28.0)	24.0	(21.0–28.0)

Most study women (87.0%) filled an outpatient opioid prescription before day 5 after cesarean delivery. Oxycodone was the most commonly filled opioid prescription (69.0% of all prescriptions), followed by hydrocodone (26.8%) and codeine (2.3%). The frequency of the dose of the postpartum opioid prescription was as follows: no opioid (13.0%), <150 total MME (10.7%), 150 to 224 MME (23.9%), 225 MME (33.7%), and >225 MME (18.8%). The proportion of women identified as White was higher in the 2 highest dose categories than in the lower dose categories. In addition, there were geographic differences in prescribing patterns across the regions of Tennessee. A larger percentage of women that did not fill a postpartum opioid prescription had evidence of severe maternal morbidity (3.2%) compared with women that filled a postpartum opioid prescription at any dosage (range, 2.3%–2.8%). Among women that filled any postpartum opioid prescription, we noted higher proportions with baseline tobacco use, depression medication use, benzodiazepine use, and >1 outpatient visits during pregnancy (Table 1).

The risk of serious opioid-related events

We identified 1637 SOREs during the study period. The rate of SOREs was 3.0 per 100 person-years. The earliest event identified for most patients experiencing a SORE was the development of persistent opioid use (1.7% of women), followed by new OUD diagnosis or methadone or buprenorphine fill (0.7%), opioid-related overdose (0.2%), and opioid-related death (n=4 [0.0%]). Nearly two-thirds of the population reached the end of a 1-year follow-up (66.3%), 31.1% lost enrollment, and 0.1% died from non–opioid-related causes before the end of 1 year after delivery (Table 2).

TABLE 2

Reasons for end of follow-up among women with cesarean delivery stratified by overall opioid morphine milligram equivalent quartiles of the first postpartum opioid prescription, Tennessee Medicaid (2007–2014)

			Total MME of the first postpartum opioid prescription
	Total n=65,170		0 n=8447		1–<150 MME n=6948		150–<225 MME n=15,572		225 MME n=21,953		>225 MME n=12,250
	n	%	n	%	n	%	n	%	n	%	n	%
Reason for end of follow-up
Persistent use	1136	1.7	73	0.9	107	1.5	256	1.6	417	1.9	283	2.3
Methadone-buprenorphine fill	162	0.3	15	0.2	7	0.1	32	0.2	74	0.3	34	0.3
Opioid use disorder diagnosis	300	0.5	37	0.4	24	0.3	50	0.3	128	0.6	61	0.5
Opioid-related overdose	35	0.1	6	0.1	7	0.1	7	0	11	0.1	4	0
Opioid-related death	4	0.0	1	0	1	0	1	0	1	0	0	0
Death	42	0.1	5	0.1	5	0.1	15	0.1	12	0.1	5	0
Loss of enrollment	20,286	31.1	2472	29.3	2229	32.1	4828	31	6791	30.9	3966	32.4
End of follow-up	43,205	66.3	5838	69.1	4568	65.7	10,383	66.7	14,519	66.1	7897	64.5

In the primary analysis using multiple imputation and accounting for relevant covariates, we observed a nonlinear association between opioid dosage and the risk of SOREs. Compared with women that did not fill a postpartum opioid prescription, we observed a significantly elevated risk at total doses of ≥100 MME, with increasing risk as the total opioid dose increased (Figure 2, Supplemental Table 1).

Adjusted HRs for SOREs at each observed value of the dose of the first postpartum opioid prescription

The adjusted HRs for SOREs are shown at each observed value of the dosage (up to the 99th percentile of the dose distribution) of the first postpartum opioid prescription from a model with a binary indicator variable for filling a postpartum opioid prescription and with restricted cubic spline terms to account for the dosage of the first filled postpartum opioid prescription among patients that filled an opioid prescription (knots: 125, 150, 225, 275, and 600). The model accounted for baseline comorbidities, medication use, delivery complications, and multiple deliveries, Tennessee Medicaid (2007—2014). We calculated these exponentiated linear predictions while setting the values of every other covariate to their reference.

CI, confidence interval, HR, hazard ratio; MME, morphine milligram equivalent; SORE, serious opioid-related event.

Wiese. Postpartum opioid use after cesarean delivery and the risk of opioid-related harms. Am J Obstet Gynecol MFM 2021.

Secondary analyses

In a secondary analysis examining the impact of the daily opioid dose, we observed that women with a daily dosage of >21 MME per day had a significantly higher risk of SOREs than women that did not fill an opioid prescription, with the risk increasing as the daily dosage increased (Supplemental Figure 2). Except for the lowest observed doses (≤6 MME per day), women prescribed a daily dosage of ≤21 MME per day had an elevated, but not a significantly increased, risk.

Post hoc analyses examined the predicted risk of SOREs at specific values of commonly dispensed total tablet quantities and days' supply stratified by opioid type and formulation strength (hydrocodone [n=15,185] and oxycodone [n=39,136]), excluding the other less frequently used opioid types (combined, n=2402). Filling an opioid prescription with ≥3 days of supply with ≥30 tablets was consistently associated with a significantly higher risk of SOREs, regardless of opioid type or strength (Figure 3, Supplemental Table 2). We observed an elevated risk of SOREs among women that filled any number of hydrocodone tablets for any prescribed days' supply, although findings were not significant for prescriptions with 2 days' supply or <10 total tablets. We observed a nonsignificantly lower risk of SOREs among women that filled 10 oxycodone tablets (≤5 mg) regardless of the dispensed days' supply.

Adjusted hazard ratios for SOREs stratified by opioid type and formulation strength

Adjusted hazard ratios for SOREs for selected numbers of tablets stratified by opioid type and formulation strength. The model included a binary indicator for filling a postpartum opioid prescription and restricted cubic spline terms modeling the prescribed number of tablets (knots: 20, 30, and 40) and days' supply (knots: 2, 4, and 7) for each opioid type and strength. The model accounted for baseline comorbidities, medication use, delivery complications, and multiple deliveries, Tennessee Medicaid (2007–2014). We calculated the exponentiated linear predictions while setting the values of every other covariate to their reference. The "n" in each model represents the population of patients that filled the specific opioid type (reference group "no opioid," n=8447).

CI, confidence interval; SORE, serious opioid-related event

Wiese. Postpartum opioid use after cesarean delivery and the risk of opioid-related harms. Am J Obstet Gynecol MFM 2021.

Sensitivity analyses

Findings in sensitivity analyses were consistent with the primary analysis (Supplemental Figures 3-6). We observed subhazard ratios for each dose category in the competing risk survival analysis that were nearly identical to the adjusted HRs in a noncompeting risk Cox regression model.

Comment

Principal findings

Opioid-naïve women who underwent cesarean delivery and filled an outpatient postpartum opioid prescription had an elevated risk of SOREs compared with women who did not. This association was especially pronounced at prescribed total doses of >100 MME, although risks associated with lower total doses cannot be excluded. Furthermore, increased risks of SOREs were demonstrated for several commonly prescribed opioid doses and quantities, signaling an opportunity for potential improvement in prescribing.

Results in the context of what is known

In a previous study of the TennCare population, filling ≥2 opioid prescriptions in the 42 days after delivery was strongly associated with a higher risk of SOREs.⁵ In a subset of this population that underwent vaginal delivery, the risk of SOREs was significantly higher among women filling any postpartum opioid dose compared with those that did not fill an opioid prescription, demonstrating that there was no "safe" postpartum opioid dose for women after vaginal delivery.²¹ These findings align with a study among a nationwide sample of the Medicaid population that demonstrated women that filled a postpartum opioid prescription after vaginal delivery have a higher risk of becoming a chronic opioid user than those that did not fill a postpartum opioid prescription.²²

However, few studies have examined the safety of the initial postpartum opioid dose among women after cesarean delivery.²³ A previous study reported that there was no significant increased risk of persistent opioid use among commercially insured women who filled postpartum opioid prescriptions in the higher tertiles (81.0–112.5 and >112.5 MME) than in women who filled opioid prescriptions at the lowest dose tertile (<81 MME).²⁴ Similarly, in another study of opioid-naïve women undergoing cesarean delivery (n=113,213), higher total doses of the opioid prescribed during the perioperative period (150–<225 and ≥225 MME) were not significantly associated with an increased risk of new persistent opioid use compared with the combined group of women that either filled low doses of opioids (<150 MME) or no opioids.² However, neither study included a comparison group with only women that did not receive opioids during their postpartum period. In our study, opioid-naïve women that fill an outpatient postpartum opioid prescription had an elevated risk of SOREs compared with women who did not, with the most pronounced risk at prescribed total doses of >100 MME.

Clinical and research implications

Although SOREs, such as OUD, opioid-related overdose, and death, remains relatively rare among American women, these tragic, but preventable, events represent a public health challenge because of the large number of cesarean deliveries in the United States each year (>1.2 million).^4,25 Reductions in opioid prescribing after cesarean delivery could reduce the incidence of persistent opioid use and other SOREs.

Although current recommendations focus generally on consideration of the lowest dose and shortest duration of opioid therapies, initial opioid prescriptions after cesarean delivery vary widely. Survey data from 2014 to 2016 from 6 academic medical centers demonstrated that the median prescribed tablet number at the time was 20 (interquartile range, 8–30).²⁶ A recent study from 2 hospitals that implemented an initiative to reduce opioid prescribing after cesarean delivery in 2017 demonstrated that the mean and median number of tablets prescribed at discharge after cesarean delivery was 18 oxycodone tablets after the intervention (interquartile range, 16–20).²⁷ Assuming a 5-mg formulation, 18 oxycodone tablets represent a total of 135 MME, an average dose associated with a substantial risk increase, according to our estimates. Importantly, the tablet number may have been even higher in nonintervention hospitals.²⁷ A recent study from a tertiary academic medical center in Missouri reported a median prescribed total dosage of 150 MME at discharge after cesarean delivery in 2018 (equal to 20 oxycodone 5-mg tablets).²⁸ In a separate study of a tertiary academic medical center, the median prescribed total opioid dosage was 110.7 MME in 2018, even after the passage of HB 21 that restricted opioid prescriptions to a 3-day supply (or 7 days with an exemption request).²⁹ Furthermore, we observed variations in the initial prescribed opioid dose after cesarean delivery by geographic region.

Our study provides a detailed quantification of the risk of SOREs associated with commonly prescribed opioid doses after cesarean delivery reported in the recent literature. Compared with no opioid use, the risk of SOREs increased sharply with the use of opioids, and the risks associated with the lowest doses cannot be ruled out. Less permissive recommendations for opioid prescribing are warranted. The detailed quantification of risks estimated in this study has important implications for future initiatives, such as providing education in areas with high prescribing patterns, that aim to reduce opioid prescribing after cesarean delivery. Our study demonstrates that routine prescribing of opioids after cesarean delivery at commonly prescribed doses should consider the risk associated with such practice and that postpartum pain management after cesarean delivery should be a shared decision between patient and provider that considers multimodal therapy with nonopioid alternatives when appropriate.³⁰

Strengths and limitations

Our study enhances the existing research in this area by comparing women filling a postpartum opioid prescription at different doses with women not filling a postpartum opioid prescription, in contrast with a referent category that included both women filling a low opioid dose prescription and women who did not fill a postpartum opioid prescription. Combining categories of women who filled no opioid prescription and those who filled low MME prescriptions could mask important differences between these 2 groups. Furthermore, previous studies that categorized opioid exposures based on dose distributions from the aforementioned groupings might have led to the creation of artificial cut points that may hide true associations. We designed an analysis using continuous MME as the primary exposure and used women who did not fill a postpartum opioid prescription as the reference group. When analyzed continuously, our study demonstrated a consistently elevated risk of SOREs associated with any opioid dose, although estimates were less precise for the less frequent doses of <100 MME.

A major strength of our study was the linkage of administrative data with state vital records data (birth and death certificates) and a state hospitalization registry to precisely measure exposures, outcomes, and covariates of interest. However, our study was limited by reliance on outpatient filled pharmacy data as we could not characterize inpatient opioid administration or observe actual opioid consumption after discharge. Although filled pharmacy data are reliable and free of recall bias, previous work has shown that many women may use fewer than the number of opioids prescribed after cesarean delivery.^9,31-33 We addressed concerns regarding the incomplete identification of OUDs before pregnancy in our primary analysis (6 months of baseline data before delivery) by conducting a sensitivity analysis that required at least 12 months of baseline data available before delivery, which found similar results. We accounted for several covariates potentially associated with the dose of the initial postpartum opioid and the risk of SOREs and supplemented traditional claims data with birth certificate information. We did not account for covariates for which there was no evidence of an association between the variable and either the postpartum opioid dose or the risk of SOREs (ie, maternal weight).^3,31,34 However, we cannot rule out the possibility of residual confounding and the potential misclassification of covariates because of incomplete or incorrect diagnosis histories. In addition, our findings may not be generalizable to privately insured women or those with more recent cesarean delivery. Nevertheless, as the focus of the study was to identify the risk of SOREs at different doses of the initial postpartum opioid prescription compared with not filling an opioid prescription after cesarean delivery, it was beneficial to focus on a period with a higher and broader distribution of the dose of the opioid unimpacted by the change to the ICD-10 coding system and the opioid reduction initiatives implemented in Tennessee since 2016.

Conclusion

Among previously opioid-naïve women after cesarean delivery, women who filled a postpartum opioid prescription had a higher risk of SOREs than women who did not fill a postpartum opioid prescription. Current guidelines should consider the increased risk of SOREs associated with various common initial postpartum opioid doses prescribed after cesarean delivery.

​

AJOG MFM at a Glance

Why was this study conducted?

The risk of serious opioid-related events (SOREs) associated with postpartum opioid availability after cesarean delivery remains unclear. We examined whether the risk of SOREs varied based on the characteristics of the first postpartum opioid prescription after cesarean delivery.

Key findings

After cesarean delivery, women with commonly prescribed opioid amounts available had an elevated risk of SOREs compared with women without an opioid prescription. Evidence of an increased risk was limited for women receiving low daily doses and women receiving low tablet numbers at lower strength formulations.

What does this add to what is known?

Reducing the dosage of opioids prescribed after cesarean delivery can mitigate the risk of SOREs.

Supplementary Material

Supplement

ACKNOWLEDGMENTS

We are indebted to the Division of TennCare of the Tennessee Department of Finance and Administration, which provided data for the study. Moreover, we are indebted to the Tennessee Department of Health for providing data for the study.

A.D.W. was supported by the Building Interdisciplinary Research Careers in Women's Health program supported by the National Institutes of Health (NIH) (grant number 5K12HD043483) and by the National Institute on Drug Abuse (NIDA) (grant number K01DA051683). C.G.G. was supported by the NIH National Institute on Aging (grant number R01AG043471). S.S.O. was supported by the NIH (grant number K23DA047476) and S.W.P. was supported by NIDA (grant number R01DA045729). The funders did not have a role in any aspect of developing the report or in the decision to submit the article for publication.

Footnotes

C.G.G. reports receiving consulting fees from Pfizer and Merck and research support from Sanofi Pasteur and Campbell Alliance/Syneos Health. A.D.W. reports receiving consulting fees from the Tennessee Department of Health. The remaining authors report no conflict of interest.

The findings of the study were presented as an oral presentation at the annual conference of the International Society for Pharmacoepidemiology, held virtually, August 21–25, 2021.

A.D.W. had full access to the data in the study and takes responsibility for the integrity of the data and accuracy of the data analysis.

Supplementary materials

Supplementary material associated with this article can be found in the online version at doi:10.1016/j.ajogmf.2021.100475.

Contributor Information

Andrew D. Wiese, Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN.

Sarah S. Osmundson, Obstetrics and Gynecology, Vanderbilt University Medical Center, Nashville, TN.

Edward Mitchel, Jr, Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN.

Margaret Adgent, Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN.

Sharon Phillips, Biostatistics, Vanderbilt University Medical Center, Nashville, TN.

Stephen W. Patrick, Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN; The Center for Child Health Policy, Vanderbilt University Medical Center, Nashville, TN.

Andrew J. Spieker, Biostatistics, Vanderbilt University Medical Center, Nashville, TN.

Carlos G. Grijalva, Department of Health Policy, Vanderbilt University Medical Center, Nashville, TN ; Geriatric Research, Education and Clinical Center, Tennessee Valley Healthcare System, US Department of Veterans Affairs, Nashville, TN.

References

1. Wakeam E, Molina G, Shah N, et al. Variation in the cost of 5 common operations in the United States. Surgery 2017;162:592–604. [PubMed] [Google Scholar]

2. Peahl AF, Dalton VK, Montgomery JR, Lai YL, Hu HM, Waljee JF. Rates of new persistent opioid use after vaginal or cesarean birth among US women. JAMA Network Open 2019;2:e197863. [PMC free article] [PubMed] [Google Scholar]

3. Osmundson SS, Schornack LA, Grasch JL, Zuckerwise LC, Young JL, Richardson MG. Postdischarge opioid use after cesarean delivery. Obstet Gynecol 2017;130:36–41. [PubMed] [Google Scholar]

4. Osmundson SS, Min JY, Grijalva CG. Opioid prescribing after childbirth: overprescribing and chronic use. Curr Opin Obstet Gynecol 2019;31:83–9. [PMC free article] [PubMed] [Google Scholar]

5. Osmundson SS, Min JY, Wiese AD, et al. Opioid prescribing after childbirth and risk for serious opioid-related events: a cohort study. Ann Intern Med 2020;173:412–4. [PMC free article] [PubMed] [Google Scholar]

6. Osmundson SS, Wiese AD, Min JY, et al. Delivery type, opioid prescribing, and the risk of persistent opioid use after delivery. Am J Obstet Gynecol 2019;220:405–7. [PMC free article] [PubMed] [Google Scholar]

7. Badreldin N, Grobman WA, Yee LM. Inpatient opioid use after vaginal delivery. Am J Obstet Gynecol 2018;219:608.e1–7. [PubMed] [Google Scholar]

8. Badreldin N, Grobman WA, Yee LM. Association between neuraxial analgesia and postpartum opioid use following vaginal delivery. Am J Obstet Gynecol 2020;222:278–9. [PubMed] [Google Scholar]

9. Komatsu R, Carvalho B, Flood PD. Recovery after nulliparous birth: a detailed analysis of pain analgesia and recovery of function. Anesthesiology 2017;127:684–94. [PMC free article] [PubMed] [Google Scholar]

10. Peahl AF, Morgan DM, Dalton VK, et al. New persistent opioid use after acute opioid prescribing in pregnancy: a nationwide analysis. Am J Obstet Gynecol 2020;223:566.e1–13. [PMC free article] [PubMed] [Google Scholar]

11. Min JY, Osmundson SS, Griffin MR, et al. Variation in discharge opioid prescribing after vaginal births. Am J Obstet Gynecol MFM 2021;3:100298. [PMC free article] [PubMed] [Google Scholar]

12. Peahl AF, Smith R, Johnson TRB, Morgan DM, Pearlman MD. Better late than never: why obstetricians must implement enhanced recovery after cesarean. Am J Obstet Gynecol 2019;221:117.e1–7. [PubMed] [Google Scholar]

14. Ray WA, Chung CP, Murray KT, Hall K, Stein CM. Prescription of long-acting opioids and mortality in patients with chronic noncancer pain. JAMA 2016;315:2415–23. [PMC free article] [PubMed] [Google Scholar]

16. Jenkner C, Lorenz E, Becher H, Sauerbrei W. Modeling continuous covariates with a "spike" at zero: bivariate approaches. Biom J 2016;58:783–96. [PubMed] [Google Scholar]

17. Lorenz E, Jenkner C, Sauerbrei W, Becher H. Modeling variables with a spike at zero: examples and practical recommendations. Am J Epidemiol 2017;185:650–60. [PubMed] [Google Scholar]

18. Marrie RA, Dawson NV, Garland A. Quantile regression and restricted cubic splines are useful for exploring relationships between continuous variables. J Clin Epidemiol 2009;62. 511–7.e1. [PubMed] [Google Scholar]

19. Rubin DB. Multiple imputation for nonresponse in surveys. New York, NY: John Wiley & Sons; 2004. [Google Scholar]

20. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999;94:496–509. [Google Scholar]

21. Wiese AD, Osmundson SS, Mitchel E, et al. Prescription opioid dose after vaginal delivery and the risk of serious opioid-related events: a retrospective cohort study. Womens Health Issues 2021;31:376–83. [PMC free article] [PubMed] [Google Scholar]

22. Zhu Y, Huybrechts KF, Desai RJ, et al. Prescription opioid use after vaginal delivery and subsequent persistent opioid use and misuse. Am J Obstet Gynecol MFM 2021;3:100304. [PubMed] [Google Scholar]

23. National Academies of Sciences, Engineering, and Medicine. Framing opioid prescribing guidelines for acute pain: developing the evidence. Washington, DC: National Academies Press; 2020. [Google Scholar]

24. Bateman BT, Franklin JM, Bykov K, et al. Persistent opioid use following cesarean delivery: patterns and predictors among opioid-naïve women. Am J Obstet Gynecol 2016;215:353.e1–18. [PMC free article] [PubMed] [Google Scholar]

25. Ladha KS, Gagne JJ, Patorno E, et al. Opioid overdose after surgical discharge. JAMA 2018;320:502–4. [PMC free article] [PubMed] [Google Scholar]

26. Bateman BT, Cole NM, Maeda A, et al. Patterns of opioid prescription and use after cesarean delivery. Obstet Gynecol 2017;130:29–35. [PMC free article] [PubMed] [Google Scholar]

27. Landau R, Romanelli E, Daoud B, et al. Effect of a stepwise opioid-sparing analgesic protocol on in-hospital oxycodone use and discharge prescription after cesarean delivery. Reg Anesth Pain Med 2021;46:151–6. [PubMed] [Google Scholar]

28. Veade A, McKinnish T, Carter E, Lewkowitz A. Associations among discharge opioid prescribing and inpatient postpartum opioid usage after. AJP Rep 2020;10:e275–80. [PMC free article] [PubMed] [Google Scholar]

29. Potnuru PP, Patel SD, Birnbach DJ, Epstein RH, Dudaryk R. Effects of state law limiting postoperative opioid prescription in patients after cesarean delivery. Anesth Analg 2021;132:752–60. [PubMed] [Google Scholar]

30. ACOG Committee Opinion No. 742 summary: postpartum pain management. Obstet Gynecol 2018;132:252–3. [PubMed] [Google Scholar]

31. Schmidt P, Berger MB, Day L, Swenson CW. Home opioid use following cesarean delivery: how many opioid tablets should obstetricians prescribe? J Obstet Gynaecol Res 2018;44:723–9. [PMC free article] [PubMed] [Google Scholar]

32. Komatsu R, Carvalho B, Flood P. Prediction of outliers in pain, analgesia requirement, and recovery of function after childbirth: a prospective observational cohort study. Br J Anaesth 2018;121:417–26. [PMC free article] [PubMed] [Google Scholar]

33. Nakahara A, Ulrich ND, Gala R. Opiate prescribing practices in the postpartum unit. Ochsner J 2018;18:36–41. [PMC free article] [PubMed] [Google Scholar]

34. Howard R, Fry B, Gunaseelan V, et al. Association of opioid prescribing with opioid consumption after surgery in Michigan. JAMA Surg 2019;154:e184234. [PMC free article] [PubMed] [Google Scholar]

pleasantmilloon.blogspot.com

Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8599660/

Share this post