Are opioids being over-prescribed in gynaecology surgery?

Adequate analgesia is vital in the recovery of gynaecological surgical patients. Pain is commonly experienced by patients following gynaecological surgery and even minimally invasive techniques, such as laparoscopy, may lead to postoperative pain in up to 35–80% of patients (van Dijk et al, 2018). Moderate or severe post-operative pain can negatively influence factors such as mobilisation, length of hospital stay and time taken to return to usual activities of daily living (Garimella and Cellini, 2013). Uncontrolled acute pain may also lead to chronic pain. Minimally invasive gynaecological surgical techniques, such as laparoscopic hysterectomy and myomectomy, have contributed to a reduction in the duration of inpatient stay and patient recovery times (Johnson et al, 2005; Barakat et al, 2011). However, analgesia is still required for patient recovery and these requirements extend from the inpatient stay to an undefined period of time post-discharge.

It is common practice for opioids to be an analgesia choice in post-operative patients but opioid use for acute pain carries the risk of long-term opioid use (Shah et al, 2017; Brat et al, 2018). This in turn can lead to dependency, misuse and overdose. Following media-highlighted concerns regarding the volume of patients on long-term opioids in our region, a service evaluation was undertaken to identify what opioids were consumed post-operatively and to review what opioid take-home prescriptions were being prescribed to post-operative gynaecology patients.

Materials and methods

Data were retrospectively collected from patient cases that were admitted to the gynaecology ward following gynaecology surgery at a tertiary care hospital in the UK, between 1 July 2019 and 30 September 2019. Using pharmacy software, take-home prescriptions of opioids that were dispensed from the gynaecology ward during this time period were identified. In total, 253 cases were highlighted and the first 100 case notes were initially requested. The case notes and discharge summaries were then reviewed; only post-operative cases were included, and gynaecology oncology patients were subsequently excluded. The resulting number of cases for inclusion was 40. Although the intention was to request and review the case notes for the remaining 153 cases, the COVID-19 pandemic prevented this. Hospital services were prioritised, and administrative support, including case note delivery, for all but essential service evaluations was interrupted. Therefore, increasing the sample size above 40 included cases was unfortunately not possible.

Inpatient opioid consumption was reviewed, and an average daily inpatient opioid usage was calculated in each case. Four cases had an unrecorded duration of admission and therefore were not included in some aspects of the data analysis. An average oral morphine milligram equivalent (MME) dose was calculated for each case, in addition to an opioid dose equivalent which was consistent with the opioid choice for discharge. This allowed for an appreciation of whether patients were already receiving the maximum dose of the opioid which they were discharged home with (take-home opioid), prior to discharge. See Table 1 for an explanation of MME and an approximation guide for the relative potency of opioids, taken from the British National Formulary (Joint Formulary Committee, 2022).

The take-home prescriptions were reviewed to look at the choice of opioid, the dosage and the duration prescribed. Data were also collected on several points of interest relating to patient factors, surgical factors and anaesthetic factors. It was thought that these factors may influence the clinical decision of the prescription of take-home opioid prescribed. Patient factors included age, body mass index (BMI), past medical history, drug history, pre-existing opioid use, hospital opioid take-home prescription in the preceding 12 months, opioid allergies/intolerances and inpatient opioid usage. Surgical factors, such as the procedure performed and any intra-operative or post-operative complications, were recorded. Anaesthetic data collected included the ASA physical status classification score, type of anaesthetic used and whether a post-operative epidural or intravenous patient-controlled analgesia (PCA) was used.

It was standard practice on the gynaecology ward for non-pharmacological pain management to be offered to patients, although specific data collection on this was not within the scope of this evaluation. Non-pharmacological recovery techniques included post-operative explanations of the surgical procedure, emotional support from staff and patient positioning. It also included peppermint capsules, restoration and encouragement of a balanced diet and early mobilisation, with guided support by physiotherapists where needed.

A literature review was also undertaken to review the current evidence for post-operative opioid requirements in gynaecology patients and to search for any pre-existing guidance on take-home prescribing in post-operative gynaecology patients.

As the study was performed in the context of a service evaluation, ethical approval was not required.

Results

Patient case characteristics

Patients' baseline characteristics and surgical procedure list can be seen below in Table 2. Four patients had received one hospital-dispensed opioid take-home prescription in the preceding 12 months and five patients were receiving a pre-existing regular opioid prescription. None of the patients had a postoperative epidural or PCA. There was one intraoperative complication; a sigmoid bowel injury that occurred during a laparotomy case for a pelvic abscess.

Main results

The opioids prescribed to-take-home in each of the 40 cases included dihydrocodeine tartrate, codeine phosphate and tramadol hydrochloride. The most common opioid choice across all age groups was codeine phosphate (35/40) with a dosage prescription of 30–60 mg (Table 3). Most prescriptions were to be taken ‘when required’ (29/40). The majority of cases were prescribed seven or more days of opioids (27/34) and only three cases were prescribed three days or less duration (3/34). The average duration of opioid prescription was 7.9 days (range 3-28 days). One case was prescribed 28 days of opioids (total laparoscopic hysterectomy). For six prescriptions, although the dose and frequency of the codeine phosphate prescriptions were specified, the prescribers did not specify the duration of the prescription. In these cases, the pharmacy team dispensed a 28-tablet box of 30 mg codeine phosphate to take home. To note, in each patient case, a pharmacy would also dispense no more than 28 tablets of opioids. Meaning that the prescriptions of greater than seven days (n=5) would have been limited to 28 tablets being dispensed.

Once transferred to the gynaecology ward, nine cases (24%) did not consume any inpatient opioids and only three cases (8%) received the maximum daily dose of take-home opioid equivalents on average each day during admission (see Table 2). The average daily oral MME dose was greatest in the laparotomy category (17.0) and lowest in the diagnostic laparoscopy category (4.5). Non-opioid analgesia options such as paracetamol and/or ibuprofen were taken in 35/40 cases, often being used as first-line management of pain control.

Most prescribers were postgraduate doctors with two to four years of experience (37/40), two had four to seven years of experience (2/40) and one was in their first postgraduate year (1/40). These junior doctors worked for all the consultants in the gynaecology department and therefore prescribing practice was independent of the named consultant.

An unexpected finding was that the laparoscopic patient group (n=15) was prescribed on average longer durations of opioids than the laparotomy group (n=8) (8.6 vs 7.1 days). Furthermore, the laparoscopic hysterectomy group were prescribed a longer average duration of opioids than the abdominal hysterectomy group (12.6 vs 10.5 days) (see Table 4). The patient demographics, drug history and grade of the prescriber were assessed in the laparoscopic and laparotomy groups to review if there were any differences in the patient groups that may account for these findings, ie provide a clinical justification.

The average BMI was identical for both groups (BMI 28.7) and the average age of patients were similar, but older, in the laparotomy group (45.6 vs 40.5 years) (see Table 4). Each group had two patients who had received a take-home opioid prescription in the preceding 12 months. Neither group had any patients who received the maximum average daily dose of take-home opioids, whilst an inpatient, nor any patients with intolerances or allergies to analgesia. Only one of the laparoscopy cases had a pre-existing regular opioid prescription; this patient had a past medical history inclusive of arthritis and Crohn's disease, was on regular dihydrocodeine and was discharged home with codeine phosphate for 14 days. Both the laparoscopy and laparotomy groups had only one prescription completed by a specialist registrar. Therefore, overall the groups were similar in age, BMI, drug history and grade of prescriber.

Discussion

In the absence of national guidance on take-home prescriptions of analgesia in the post-operative gynaecology patient, clinicians often make independent decisions and prescribe according to their clinical knowledge, interpretation of the patient's requirements and personal preferences. Education and decision-making tools for prescribers and/or local standardised protocols for pain assessment, may support these decision-making processes and encourage safe, appropriate and patient-centred prescribing. From this service evaluation, a lack of patient-centred prescribing may be evidenced by the 21/34 patients receiving a seven-day prescription regime of opioids and by the laparoscopic group of patients receiving longer durations of opioids than the laparotomy group (8.6 vs 7.1 days). With no identifiable clinical differences between these two groups to support longer durations in the laparoscopy group, this is inappropriate prescribing practice. Recognised advantages of laparoscopic surgery are shorter hospital stays, earlier return to normal activities and reduced post-operative pain (Johnson et al, 2005; Barakat et al, 2011). As an opioid prescription, duration is recognised as the strongest predictor of future opioid misuse (Shah et al, 2017; Brat et al, 2018). Clinicians should be aiming to prescribe opioids for the shortest duration as required by the patient.

Therefore, it is important to consider how pain is assessed and then more specifically, how opioid requirements are determined. At this tertiary centre, local guidelines advocate for standardisation of prescribing practices by the anaesthetic team for post-operative analgesia during inpatient care. The objective of these guidelines is to establish consistency in take-home opioid prescriptions for patients.. Predicting how much pain relief patients will need once discharged is challenging, but it may be inferred from how much they consume. This consumption includes inpatient use and post-discharge use. This service evaluation demonstrated that the average daily oral MME dose varied between 4.5 and 17.0 according to the gynaecological surgical procedure. There was a lower inpatient opioid consumption in the laparoscopic patients compared to the laparotomy patients (10.2 vs 17.0 average daily oral MME). Yet, the laparoscopic hysterectomy group were prescribed a longer average duration of opioids than the abdominal hysterectomy group (12.6 vs 10.5 days). These results suggest that prescribing may not have been guided by consideration of the patients' inpatient consumption or that there was a lack of awareness of the evidence which shows reduced post-operative pain in this patient group.

In this tertiary centre, an Opioid Working Group had been recently established, with a purpose to address the variation of opioid prescribing which had been identified across different surgical specialties, including gynaecology. Reduced variation would improve time/administrative/cost effectiveness, as well as patient safety. The Opioid Working Group allows collaborative working between the pharmacy team, the acute pain team, anaesthetists, and the governance team. It aims to promote standardisation across different departments, and this will involve reviewing the current practice of the other surgical specialties as well as gynaecology. This working group could also allow for the delivery of training and education to prescribers which specifically reflects the resulting recommended practice. The group's proposed strategy involves reviewing patient consumption of opioids post-discharge, which would allow an opportunity to collect information on the patient's experience and feedback which is currently missing and provide evidence for usage according to surgical procedure and quantify wastage.

The creation of a standard prescription that advises which opioid and how much, should be based upon a predetermined evidence-based analysis of patient need and should recommend avoiding opioid prescriptions where appropriate. This could also avoid significant wastage of medication, as in this service evaluation, 24% of patients did not take any opioid analgesia post-operatively whilst admitted on the gynaecology ward and yet were still discharged home with take-home opioids. Clinicians may be concerned that prescribing too little analgesia may result in patients experiencing uncontrolled pain, but in these cases these patients had opioids unnecessarily prescribed. These results are supported by Thiels et al (2018) prospective multicentre study, which reviewed the post-operative opioid consumption following the discharge of 139 gynaecology patients who had undergone minimally-invasive hysterectomy. In this study, one-third of surgical patients consumed no opioids post-discharge. These findings should be used to support the opinion that patients require less take-home opioid analgesia than the clinician prescribers in this service evaluation assumed.

The results from this service evaluation have highlighted that education and training would be of significant benefit to clinicians' prescribing practice. No standardised local protocol or guidance for take-home opioid prescribing in gynaecology patients was available and there was no regularly timetabled teaching, which focussed on the assessment of pain and rationalised decision-making of take-home prescribing for clinicians in this department. Lack of clinical knowledge, awareness of current literature, insight into opioid dependency risk or lack of local and national guidance therefore may be factors that contributed towards the overprescribing of take-home opioids. In addition to formalised teaching, directing prescribers to other sources of support would have been useful; this may include the pain team or other senior members of the surgical team. Educating clinicians regarding the risk of over-prescribing opioids has been shown to result in reduced prescription durations (Evans et al, 2020) and education targeted at those most common prescribers of take-home opioids, such as junior doctors in this case, would offer the most benefit. There were numerous opioids prescribed in this service evaluation (dihydrocodeine, codeine and tramadol) and consideration could be made, through training and education, to promote the familiarisation of a single opioid to support appropriate prescribing. The importance of completing all components of a prescription should also be emphasised in teaching, as in six prescriptions in this service evaluation, the prescribers did not specify the duration of the codeine prescription.

Reducing the duration of take-home opioids by suggesting a limit could be the key step to reducing opioid dependency risk. Although the pharmacy dispensing limit of 28 tablets is a method of avoiding patients taking home more than 7 days of regular opioids, this has scope to be reduced further. The optimal limit should also aim to avoid poorly controlled pain and avoid the need for patients to re-present requesting additional pain relief. In the study looking at opioid prescribing in 605 post-operative gynaecology patients (Mark et al 2018), the duration of take-home opioids was restricted to three days in both the laparoscopy and laparotomy groups. This did not significantly increase the number of refill requests or post-operative pain scores, providing evidence of sufficiency. The opioids used in this American study were hydrocodone and oxycodone, which are less commonly prescribed as outpatient prescriptions in the UK. However, when calculating the corresponding opioid dosing regimen with dihydrocodeine, codeine phosphate or tramadol, it allows for an appropriate equivalent in all dosing regimens, except their highest opioid regime of oxycodone 10 mg QDS. In these circumstances, UK patients are usually not discharged home when still using inpatient-strong opioids such as morphine, so patients would not be expected to go home with that ongoing level of opioid requirement. A similar level of opioid requirement is also supported by the Overton et al (2018) study, which used a Delphi method of expert opinion, to recommend between 0-10 or 0-20 5mg oxycodone tablets depending on gynaecological procedure. Therefore, these results in conjunction with the majority (92%) of patients in this service evaluation not consuming the maximum daily dose of take-home opioids prior to discharge, would provide supportive evidence that a proposed limit of three days of opioids would be sufficient for post-operative gynaecology patients.

Conclusion

Patients undergoing minor or major gynaecological surgical procedures are at risk of developing new persistent opioid use (Brummett et al, 2017) with up to 6% of patients continuing to use opioids 3 months post-operatively. This is significant in this tertiary centre, which performs approximately 2300 non-oncological gynaecology procedures each year. Guidelines, both local and national, can allow standardised recommendations, are educational and can result in clinicians adjusting their practice of prescribing accordingly. There is currently no national UK guidance for take-home analgesia in post-operative patients.

Supported by evidence from current literature and the results collected from this service evaluation, if guidance could recommend a limit of 3 days of take-home opioids, this could be the most direct way to change prescribing practice with maximum risk reduction. In this service evaluation only three cases (9%), were prescribed a 3-day duration of opioids or less. This suggests there is a large scope for improvement in reducing durations of opioid prescriptions without adverse pain-related outcomes. Supporting prescribers via training and education could also avoid a significant proportion of patients (24% in this service evaluation) from being prescribed unnecessary opioid prescriptions when they have not consumed any whilst an inpatient.

The authors would recommend that the next step is to undertake a data collection looking at opioid consumption post-discharge in gynaecology patients who have undergone a greater range of procedures including vaginal surgery, diagnostic and operative laparoscopy and laparotomies for indications in addition to hysterectomy, as this evidence is lacking in the current literature. The results could be used to develop guidance which could hopefully guide procedure-specific discharge prescriptions. Furthermore, the development and implementation of pain assessment and decision-making tools based on this guidance, could enable junior staff to make evidence-based, rational choices promoting effectiveness, as well as patient safety.