Evaluation of a new departmental policy to decrease routine opioid prescribing after vaginal delivery

Abstract

BACKGROUND:

In line with a nationwide commitment to decrease opioid prescribing, in October 2017, our department implemented a new departmental policy to cease routine provision of opioid prescriptions at the time of discharge following vaginal delivery.

OBJECTIVE:

This study aimed to evaluate the effect of this policy on the number of discharge opioid prescriptions provided and outpatient encounters observed postpartum.

STUDY DESIGN:

This was a retrospective cohort study of patients who underwent vaginal delivery at our institution from November 2016 to January 2018. We reviewed delivery and postpartum records for patients 18 years or older, without chronic opioid use or contraindication to nonsteroidal antiinflammatory medication use. The primary outcome was the proportion of patients provided with an opioid prescription at the time of discharge following vaginal delivery. The secondary outcome was the number of unscheduled patient encounters related to pain in the 6-week postpartum period. Fisher’s exact test was used to compare these outcomes before and after implementation of the new departmental opioid-prescribing policy.

RESULTS:

A total of 1188 charts were reviewed; among those charts, 810 met the inclusion criteria. Notably, 405 patients delivered before the guideline, and 405 patients delivered after its implementation. After the implementation of the new departmental policy, there was a 10-fold decrease in opioid prescriptions provided from 323 (79.8%) to 29 (7.2%) (P<.01). Although the number of unscheduled outpatient encounters postpartum increased slightly from 22 to 37 encounters after the implementation of the new departmental policy, this difference was not statistically significant (P=.08).

CONCLUSION:

Limiting opioid prescribing after vaginal delivery is associated with a considerable decrease in the number of discharge opioid prescriptions provided and does not significantly increase the number of outpatient encounters related to pain postpartum.

Introduction

Recently, there have been several published studies investigating strategies to reduce opioid use and opioid prescriptions after cesarean deliveries.^1–5 However, less attention has been given to strategies to mitigate potentially excessive opioid prescribing after vaginal deliveries.

One potential reason for providers choosing to provide opioid prescriptions after procedures is a concern for increased burden on the system with more frequent patient encounters or decreases in patient satisfaction.⁶ This concern may be impeding systems to consider adaptation of the recently published draft opioid-prescribing guidelines for uncomplicated normal spontaneous vaginal delivery that call for decreased opioid use.⁷ At least 1 study looked at this question for cesarean deliveries and found that eliminating routine ordering and prescribing of oral opioids after cesarean delivery did not change levels of pain control or patient satisfaction.⁸ However, to date, this question has not been answered for opioid prescriptions after vaginal delivery.

At Virginia Commonwealth University Health Systems (VCUHS), common practice had been to prescribe a limited number of opioid tablets after every vaginal delivery (90 morphine milligram equivalents [MME]) regardless of presence of lacerations or other potentially pain-inducing procedures. In October 2017, the Department of Obstetrics and Gynecology at VCUHS issued guidelines to discontinue routine opioid prescribing after vaginal deliveries. This study aimed todetermine what effectthis newguideline had on the number of opioid prescriptions provided at discharge after vaginal deliveries and to investigate whether there was an increase in unscheduled postpartum outpatient encounters related to pain following vaginal delivery.

Materials and Methods

This is a retrospective cohort study investigating the number of patients who were given an opioid prescription following vaginal delivery before and after implementation of the new departmental policy. The policy was developed by departmental leadership in response to published evidence and recommendations from governing bodies emerging in the opioid crisis, highlighting the probable nonindication for opioids following vaginal delivery and its eradication as a tool to decrease unnecessary opioid prescribing. The study was conducted at a single institution, Virginia Commonwealth University Hospital (VCU), a tertiary care center with approximately 2500 deliveries per year. There are 24 residents in the residency program, and all of the faculty are VCU employed. There are also certified nurse-midwives who work alongside the physicians. Most postpartum prescriptions are written by either the residents or nurses-midwives.

Three months of preinitiative data were collected from Nov. 1, 2016, to Jan. 31, 2017. This period was chosen to correlate with the period within the calendar year after the policy implementation. In October 2017, the new departmental policy was implemented, and all staff, residents, and nurses were informed regarding the policy change.

Before implementation of the new departmental policy, there were no guidelines for discharge medication orders following vaginal delivery. Patients were discharged home with varying volumes of opioid medications per provider preference. When the new departmental policy was implemented by the leadership of the department, clinicians were asked not to order or prescribe routine opioids following vaginal delivery but to continue other nonopioid prescriptions per clinical judgment. This was communicated to the residents and the faculty at the departmental meeting; no open forum discussion of this policy implementation was held, and no additional training to the staff, faculty, or residents was provided. The change of policy was not specifically communicated to the patients prenatally but was clarified to patients postpartum if questioned. No specific instructions were given regarding the opioid prescriptions for patients with vaginal deliveries complicated by lacerations or accompanied by additional procedures (ie, dilation and curettage for retained products of conception, postpartum bilateral tubal ligation).

Patients were included in the study population if they delivered vaginally during either the new departmental policy preimplementation (November 2016–January 2017) or postimplementation (November 2017–January 2018) periods with a washout period from February 2017 to October 2017 to account for a potentially slow uptake of the new guidelines. Patients were excluded if they were younger than 18 years, had an active opioid prescription at the time of admission as listed in the medication list (eg, buprenorphine or methadone for opioid-use disorder), or had an allergy or contraindication to nonsteroidal antiinflammatory drugs (NSAIDs), such as history of gastric bypass or any other medical indications precluding use of NSAIDs. Additional information on the study population was abstracted from the electronic medical record (EMR) including age, date of delivery, history of mental illness (as defined by the International Classification of Diseases, 10th revision [ICD-10] code), smoking status, type of delivery (spontaneous vaginal delivery, forceps-assisted vaginal delivery, vacuum-assisted vaginal delivery, episiotomy), additional inpatient procedures following delivery (none, bilateral tubal ligation, dilation and curettage, other), lacerations (none, first, second, third, fourth), and type of insurance (public, private, none, unknown).

The primary outcome was the proportion of patients who were given an opioid prescription at the time of discharge. We also describe the quantity of opioids provided to an inpatient and prescribed at discharge in MME before and after the implementation of the new departmental policy. The secondary outcome was the number of outpatient encounters related to pain. These encounters included visits to the emergency department, unscheduled clinic visits, or patient phone calls to the triage line within the Virginia Commonwealth University Health System. Both outcomes were abstracted from the EMR.

The primary and secondary outcomes were compared between the new departmental policy preimplementation (November 2016–January 2017) and postimplementation (November 2017–January 2018) groups using Fischer’s exact test. A P value of <.05 was used to indicate statistical significance. Wilcoxon rank-sum test was used to compare continuous variables between the groups for descriptive analyses. This study was approved by the VCU institutional review board. All analyses were performed in Stata version 10.0 (StataCorp, College Station, TX).

Results

Of the 1188 women who delivered during the study time frame, 810 met inclusion criteria and were included in the analysis (Figure 1). Half of the study sample (n=405) delivered during the new departmental policy preimplementation period, and the remaining 405 women delivered during the postimplementation period. Demographic and delivery characteristics did not differ between pre- and postimplementation periods as reported in Table 1. Mean maternal age was 29 years in both periods, and the majority were nonsmokers and did not have history of mental illness or nonopioid drug use. Insurance status was also not significantly different between the 2 periods; a third of patients each had private insurance, public insurance, and no insurance. Lacerations were more common in the new departmental policy postimplementation period (n=217) than in preimplementation (n=183) period (P=.01), but the prevalence of third- or fourth-degree lacerations did not differ between periods (P=.39; data not indicated). Postpartum procedures such as bilateral tubal ligation and immediate postpartum long-acting reversible contraception placement were more common in the postimplementation period (n=27) than in the preimplementation period (n=14) (P=.04; Table 1).

FIGURE 1.

Flowchart of patients included in the new departmental opioid-prescribing policy analysis before and after its implementation

NSAID, nonsteroidal antiinflammatory drug; VCU, Virginia Commonwealth University.

Olsen et al. Decreasing opioids after vaginal delivery. AJOG MFM 2020.

TABLE 1.

Study population characteristics before and after implementation of a new departmental opioid-prescribing policy

Characteristic	Preimplementation (n=405)	Postimplementation (n=405)	P value
Mean maternal age, y (95% CI)	28.86 (28.3–29.4)	29.16 (28.6–29.7)	.47
Insurancea			.40
Private	122	130
Public	131	142
Uninsured	151	133
Current tobacco useb			.35
Yes	29	36
No	375	366
History of mental illnessc			.25
Yes	35	45
No	368	360
Nonopioid drug use by self-reportd			.85
No	388	387
Yes	14	15
Cannabis	12	14
Cocaine	2	3
Psychostimulants	1	0
Other	1	2
Lacerations			.01
None	222	188
First degree	94	91
Second degree	76	117
Third degree	12	9
Fourth degree	1	0
Other procedures during delivery admission			.04
None	391	378
Yes	14	27
Bilateral tubal ligation	8	11
Dilation and curettage	4	3
Placement of long-acting reversible contraception:
Intrauterine device	3	9
Subdermal implant	0	3
Other	0	2

CI, confidence interval.

^aMissing 1 value;

^bMissing 4 values;

^cMissing 2 values;

^dMissing 6 values.

Olsen et al. Decreasing opioids after vaginal delivery. AJOG MFM 2020.

There was a significant reduction in the number of women provided with an opioid prescription after implementation of the new departmental policy from 323 (79.8%) to 29 (7.2%) (P<.01) (Figure 2). Patients who did not receive opioid prescriptions before the implementation of the new departmental policy were mostly patients of certified nurse-midwives; after the change in the departmental policy, patients who did not receive opioid prescriptions included patients from both the midwife and physician groups. Before the implementation, the average age of women who received opioid prescriptions was 28.44 years, 9.2% had history of mental illness, 8.9% were smoking, and 3.4% had history of nonopioid drug use. After the implementation, the average age of women who received opioid prescriptions was 31.6 years, 6.8% had history of mental illness, 17.2% were smoking, and none had history of nonopioid drug use (data not indicated). The quantity of opioids prescribed also significantly decreased from 113.5 MME (95% confidence interval [CI], 103.9–123.1) before implementation to 75.5 MME (95% CI, 61.2–89.8) after implementation of the new departmental policy (P<.01). The majority of patients who received opioid prescriptions after implementation of the new departmental policy had postpartum bilateral tubal ligations or lacerations following forceps-assisted vaginal delivery. There was not a single provider who provided most of the opioid prescriptions; however, most prescriptions were provided by the residents. In addition, inpatient opioid use decreased. Before the new departmental policy, 190 patients (46.9%) received opioids (mean MME, 37.5; 95% CI, 32.36–42.6), whereas 104 women (25.6%) received opioid after the implementation of the new departmental policy (mean MME, 29.4; 95% CI, 24.5–34.3) (P<.01).

FIGURE 2.

Opioid-prescribing patterns before and after implementation of a new departmental opioid-prescribing policy

Before the implementation of the new departmental policy, 79.8% of patients were provided with an opioid prescription (mean MME, 113.5; 95% CI, 103.9–123.1) compared with 7% (mean MME, 75.5; 95% CI, 61.2–89.8) after the implementation (P<.01).

CI, confidence interval; MME, morphine milligram equivalents.

Olsen et al. Decreasing opioids after vaginal delivery. AJOG MFM 2020.

There was some increase in postdischarge unscheduled outpatient encounters related to pain after the implementation of the new departmental policy. Before implementation, there were 22 encounters compared with 37 encounters after implementation (Figure 3). However, this difference was not significant (P=.08). No particular diagnosis or procedure accounted for the majority of pain-related encounters because the numbers were too small to analyze by group (Table 2). Only 4 of 37 postimplementation period encounters resulted in prescription of opioids; all of these encounters were because of perineal pain caused by laceration or hematoma.

FIGURE 3.

Number of postpartum outpatient encounters related to pain before and after implementation of a new departmental opioid-prescribing policy

Although 16 patients had 22 encounters before the implementation of the new departmental policy, 27 patients had 37 encounters after implementation (P values of .09 and .08, respectively).

Olsen et al. Decreasing opioids after vaginal delivery. AJOG MFM 2020.

TABLE 2.

Postpartum outpatient encounters before and after implementation of a new departmental opioid-prescribing policy

Reason for patient encounter	Preimplementation (n=16)	Postimplementation (n=27)
Pain because of vaginal laceration	4 Second-degree laceration (3) First-degree laceration with dysuria (1)	12 First-degree laceration (4) Second-degree laceration (1) Second-degree laceration cellulitis at repair site (4) Periurethral laceration (1) Perineal lesion (1) Episiotomy (1)
Pain because of operative vaginal delivery	1 Forceps-assisted vaginal delivery	0
Pelvic pain	6 Pelvic discomfort and headache (2) Cramping and vaginal pain (1) Pubic symphysis pain (1) Right lower quadrant pain (1) Pain “near ovary” (1)	5 Cramping (1) Pelvic soreness (1) Lower abdominal pain or back pain (1) Pelvic pressure (1) Pubic symphysis pain (1)
Abdominal pain (not pelvic)	1 Upper abdomen sore	1 Gallbladder pain resulting in cholecystectomy
Other	4 Bilateral nephrolithiasis (1) Unknown request for refill (1) Endometritis (1) Mastitis (1)	10 Dysuria (5) Hemorrhoids (1) Endometritis (1) Labial hematoma (1) Back pain (1) Unspecified pain (1)

Olsen et al. Decreasing opioids after vaginal delivery. AJOG MFM 2020.

Discussion

Principal findings

Following the implementation of the new departmental policy to cease routine provision of opioid prescriptions at discharge following vaginal delivery at a single institution, the proportion of women given an opioid prescription at discharge decreased 10-fold. Despite this change, the number of postdischarge outpatient encounters related to pain did not significantly increase. In line with prior work focused on cesarean delivery,⁸ our findings indicated that provider opioid guidelines for pain management after vaginal delivery are effective at decreasing opioid administration and prescribing without increasing health system burden owing to unmet pain concerns.

Results

During recent years, reports of differing strategies to decrease opioid use for acute pain (ie, postoperative period) have been published, such as shared decision-making models for opioid prescriptions^8,9 and routine use of specialized inpatient order sets.^5,10 Specifically, for vaginal deliveries, recent work has highlighted a successful initiative to decrease inpatient opioid use.¹¹ Our findings add to this existing literature indicating that system-wide policies aimed at reducing both opioid administration and prescribing are effective at changing provider actions¹² for pain management after vaginal delivery even beyond inpatient use.

Eliminating routine provision and prescribing of opioids following vaginal delivery is an area with substantial potential for combating the opioid crisis for 2 reasons. First, delivery is the most common indication for hospital admission among women of reproductive age.¹³ Second, there is much room for improvement in opioid provision practices after vaginal deliveries. A large analysis of Medicaid patients found that 12% of women filled an opioid prescription within 5 days of vaginal delivery, and of those, 1filled a second prescription 6 to 60 days after delivery, raising concern for persistent opioid use following vaginal delivery.¹⁴ This is alarming given that 78% of patients receive opioids during hospitalization for uncomplicated vaginal delivery and 29.8% receive them the day of discharge.⁷

We did not measure changes in reported patient satisfaction or reported pain severity seen with the significant decrease in opioid provision. However, our finding of a lack of change in the number of patient encounters related to pain after discharge reflect what has been seen in previous work, indicating no compromise of patient satisfaction with implementation of similar policies.¹² This finding is not surprising given the evidence that opioids are not necessary to achieve sufficient pain control for most women undergoing vaginal delivery.^15,16 In the rare circumstance that opioids are indicated, median time of use is only 1 day.¹⁵ Furthermore, it is notable that we still observed a substantial decrease in the number of opioid prescriptions provided without a significant increase in patient encounters for pain after discharge despite more patients undergoing additional painful procedures (ie, third- or fourth-degree laceration repair, dilation and curettage, bilateral tubal ligation) in the postimplementation phase.

Clinical implications

Results of this study can guide clinicians in decisions regarding prescription of opioids after vaginal delivery. It adds to the growing body of evidence that routine prescriptions of opioids postpartum can be avoided without additional burden on the health system.

Research Implications

We did not specifically measure patient satisfaction before and after the implementation of the new departmental policy. Collecting these data would provide more of a complete picture of the implications of this change.

Strengths and Limitations

This study has several limitations. First, this was a retrospective study and is subject to all the biases that are inherent to such type of study. Second, the sample size was relatively small. Data were only collected for 3 months before and after implementation of this initiative. Extending the study to review data for a year before and after implementation could provide more robust results. Data were also obtained from a single academic institution. Patients who presented to other hospitals postpartum would not have been captured in our data. Furthermore, our findings cannot necessarily be generalized to other healthcare settings. Further exploration of these outcomes in other care settings would be beneficial. Finally, all data were abstracted from the medical records rather than from measures collected for research purposes, which can introduce information bias. A strength of our study is the patient population from which the study population was abstracted, which was quite diverse including non–English-speaking patients and women across socioeconomic statuses.

Conclusion

Health systems can play important roles in the nationwide call to combat the opioid crisis. Specifically, providers can change their opioid administration and prescribing patterns with departmental leadership and support.¹⁰ As witnessed in other areas of medicine, taking such steps for pain control after vaginal delivery is feasible and does not lead to increased provider burden owing to outpatient encounters after discharge. Considering that over 2 million women are hospitalized yearly for delivery,¹⁷ women’s health providers are in a prime position to substantially combat the opioid crisis and should do so by systematically eliminating routine use of opioids following vaginal delivery.

AJOG MFM at a Glance.

Why was this study conducted?

This study aimed to evaluate the effect of the new departmental guideline on the number of discharge opioid prescriptions provided and outpatient encounters related to pain observed postpartum.

Key findings

After the implementation of the new departmental policy, there was a 10-fold decrease in opioid prescriptions provided from 323 (79.8%) to 29 (7.2%) (P<.01). Although the number of unscheduled postpartum outpatient encounters increased slightly from 22 to 37 encounters after the implementation of the new departmental policy, this difference was not statistically significant (P=.08).

What does this add to what is known?

This study adds to the growing number of studies on the use of opioids in the postpartum period, specifically focusing at the prescription patterns after vaginal delivery.