OST risk categorisation
Our exploratory assessment of hospital critical medicine lists illustrated inconsistencies in OST classification. Of the five available through a public website search, one did not include any medicines for substance dependence or analgesic opioids, one included analgesic opioids only, two specifically included methadone and buprenorphine and one listed ‘drugs for opioid dependence’ as critical in regard to omission but not delay. To understand this inconsistency, we reviewed the ODMT which, at that time, informed local critical medicines lists. Medications were deemed critical if classed as high risk (red) for either delay (‘dose not given at the time prescribed’ OR ‘dose not given within 2 h of the time prescribed’) or omission (‘not administered by the time of the next scheduled dose’). Only drugs for alcohol or opioid dependence were considered under ‘Drugs used in substance dependence’. Unlike analgesics for ‘severe chronic pain and breakthrough pain’ (Fig. 1) drugs for substance dependence were categorised as low risk if delayed and medium risk if omitted (Fig. 2).
Trust responses and policy characteristics
We identified 224 NHS trusts in England in January 2020, of which 135 met our inclusion criteria. Of these 135 trusts, 86 (64%) provided one or more relevant policies (including duplicates from two pairs of trusts with joint policies), 44 (33%) responded without a relevant policy and five (4%) did not send a definitive response, including to follow-up requests (Fig. 3). Seven trusts stated they used policies provided by local mental health trusts, but these were not provided.
Our analysis includes 101 polices from 86 trusts. Of the 86 included trusts, 35 (14%) only provided policies related to pain, peri-operative management or substance use in pregnancy (with sections on withdrawal) and five (6%) only provided general medicine policies with sub-sections on OST.
In most cases, authors of policies were listed, usually including a pharmacist and sometimes clinicians from community or liaison drug treatment services, anaesthetists, acute and emergency medicine physicians, psychiatrists, and midwives. Of the 101 relevant policies, 32 (32%) stated community drug treatment services were consulted in their development. None indicated patient consultation. Twenty-seven (28%) policies were overdue review, and 14 (14%) did not include enough information to ascertain whether this was the case.
Policy review findings
Below, we present findings under five headings. The first three are processes including barriers to timely OST. The latter two highlight general procedural barriers to OST prescription and issues of policy comprehensiveness, clarity, and tone. We forefront sections with relevant national guidance to contextualise findings.
Continuation of community prescriptions
UK Guidelines on Clinical Management for Drug Misuse and Dependence specify when people on community OST are admitted to hospital, OST prescription information is required as ‘a matter of high priority’. Prior to prescribing, an assessment is required: ‘where possible ascertaining independently when the last prescribed dose of OST was dispensed and, if possible, when it was consumed’ [27].
All policies reviewed stressed the need to independently confirm patients’ doses of community OST prior to prescribing. The means of verification was not always specified, but usually included contacting the prescriber (drug treatment service or general practitioner) and/or the dispenser (community pharmacist), with some policies requiring confirmation of when doses were last collected and if consumption was supervised. Of the 86 trusts, 18 (21%) provided alternative verification options such as contacting key workers or drug liaison services, online databases, previous hospital notes, drug charts or discharge summaries. Three trusts required written confirmation of doses from the prescriber or dispenser. Only one policy included an option of patient verification, while some explicitly stated that patients are not a trustworthy source for dose confirmation.
Verifying OST doses was often restricted to pharmacy or community drug treatment opening hours. Of the 86 trusts, 17 (20%) provided no guidance for cases when verification was not possible, and six (7%) explicitly stated that no OST should be prescribed in these circumstances. Regardless of whether community prescriptions were confirmed, some policies advised gradually increasing (titrating) doses ‘when in doubt’ or stated a lower initial dose was ‘still reasonable’. Thirty-three (38%) trusts recommended re-titration if a certain period of time had elapsed since the previous dose (normally 72 h, with a range of 48–120), often with the same regimens used for OST-naïve patients, and six recommended re-titration if consumption in the community was not supervised or if recent doses had been missed. Most guidance pertained to methadone, with only 37 (43%) trusts providing guidance on continuation of buprenorphine prescriptions. Notably, buprenorphine was not mentioned at all by 19 (22%) trusts.
New or unconfirmed prescriptions
National guidance discusses OST initiation in hospital with the proviso that ‘for patients not on OST, or where there is uncertainty about recent compliance, it is appropriate to exercise particular care’ [27]. This entails prescribing methadone in small divided doses, and titrating against opioid withdrawal symptoms, with an initial dose of no more than 10mg four times a day. The reviewed policies provided inconsistent guidance on how to manage withdrawal for patients who were OST-naïve or whose community prescription could not be confirmed.
Of the 86 trusts, 67 (78%) described a regimen for OST initiation in hospital. Others suggested this was possible but included no practical information on how to do so, sometimes advising referral to a specific hospital team or community drug service. Some did not mention that OST initiation was an option, and five (6%) prohibited it. Some trusts advised that non-opioid medications, such as anti-emetics, should be prescribed in lieu of or before OST for symptomatic relief and a few advised treating opioid withdrawal with oral morphine, codeine phosphate, or dihydrocodeine instead of methadone or buprenorphine.
Methadone regimens recommended for new or unconfirmed prescriptions included regular small (statim) doses, as required (pro re nata) regimes or various induction regimens to titrate to a regular dose. Twelve (14%) trusts stated no more than 30mg of methadone should be administered on day one (less than national guidance), and two stated no more than 20mg. Some policies did not allow for methadone doses to subsequently increase, and others limited dose increments to 5–10mg over 72 h. Only 23 (27%) trusts provided the option of initiating buprenorphine. Guidance for buprenorphine titration was not always provided; where stated, this was inconsistent between trusts, with some requiring specialist team input, which was not required for methadone initiation.
Fifty-five (64%) trusts required clinical symptoms of opioid withdrawal such as tachycardia, sweating or tremor before OST could be initiated, with some requiring specific scores on the Clinical Opiate Withdrawal Scale. Only five (6%) policies stated OST could be initiated for subjective symptoms, and most policies did not mention the psychological impact of withdrawal, which patients might face before clinical symptoms are observed. Many policies stated that patients do not develop withdrawal symptoms from methadone for 24 h and advised against prescribing any OST unless patients were admitted, sometimes for at least 24 and in one case, 48 h.
Discharge
Hospital discharge can interrupt OST if community services are not notified in advance. This is a particular risk if patients are discharged late on Friday or during the weekend. National guidelines highlight admission to hospital as an opportunity to engage people with community drug treatment services, emphasising the need to ensure that community prescriptions are restarted and prescriptions commenced in hospital are continued on discharge [27]. Despite this, 11 (13%) trusts did not include information on restarting community prescriptions post-discharge, and 26 (27%) of the 78 trusts that indicated OST could be initiated in hospital did not include information on facilitating continuation of new OST prescriptions on discharge. One policy explicitly stated hospital clinicians should not arrange continuation of new prescriptions requiring patients to self-refer to community services.
Fifty-seven (66%) trusts advised the hospital pharmacy could provide OST doses for a patient to take away in some circumstances. This often needed to be agreed with community drug services, which may not be possible if discharge occurred out of service hours. The remaining 29 (34%) trusts either did not mention providing OST to take away or explicitly stated it should not be. Most policies recommended community drug treatment services should be notified when their clients, or other patients with drug dependency were admitted to hospital and discharged; however, this was not always the case and advice was not always prescriptive.
Procedural barriers
Procedural barriers included drug test requirements and specifications for particular staff, combinations of staff or specialist teams to assess patients, confirm community doses and/or write OST prescriptions. Thirty-two (37%) trusts had policies referring to drug liaison teams in their hospitals. While liaison teams offer many benefits, these were potentially undermined by protocols requiring their presence for OST prescription and withdrawal management, posing barriers if teams were poorly resourced or operated in limited hours.
Urine drug testing is mentioned in national guidance as one form of verification prior to continuation of community OST prescriptions in hospital [27]. This is however, provided as an option rather than mandatory. Of the 86 trusts, 62 (72%) recommended drug tests for opioid dependent patients, at least in some situations. These references mainly referred to urine drug screens (UDS) and in some cases oral swabs. Some trusts required a laboratory test result positive for opioid substitution medications or opioids before OST could be prescribed rather than a point-of-care test. Of the 86 trusts, 14 (16%) required a positive UDS prior to OST prescription, regardless of whether a community prescription was confirmed; 13 (15%) required a positive test for unconfirmed or new prescriptions; and 15 (17%) suggested it was preferable to have a positive test prior to prescription. Of the 42 trusts which required or recommended positive UDS prior to OST prescription, only 10 (24%) stated point of care tests were potentially available and 19 (45%) required laboratory testing. In some cases, policies highlighted local laboratories could not process UDS out of office hours, so samples needed to be couriered to other hospitals. One trust, which required a positive UDS prior to initiating OST stated it could take up to 2 weeks to receive test results (it was not clear if point of care tests were also available). Few policies highlighted limitations of UDS; only five mentioned the possibility of false negative results, and many did not give the time frames following drug use in which test results would be the most accurate.
Comprehensiveness, clarity, and tone
Policy comprehensiveness and length varied considerably, ranging from 1 to 72 pages. Of the 86 trusts, 55 (64%) provided care pathway flow charts; however, these varied in detail and clarity. Some of the language used was unclear, for example, referring to ‘detoxification’ when the guidance related to maintenance therapy. Some policies specifically highlighted detoxification was not recommended in an acute setting because of the increased risk of overdose following discharge; however, many recommended measures which would lead to reduced tolerance by withholding OST.
Statements such as ‘opioid withdrawal is not a life-threatening condition, but opioid toxicity is’ were common. Most policies framed considerations about OST in terms of safety. These, however, often focused on verifying patients’ usual OST dose to prevent overdoses, as opposed to other considerations such as potential interactions between medicines, the possibility of increased opioid tolerance related to heroin use alongside OST in the community, or preventing premature discharges. Of the 86 trusts, 32 (37%) provided no information on contraindications and OST drug interactions. Only 42 (56%) provided policies that referred to opioid overdose management.
Many policies advised measures which appeared mistrustful of patients. Of the 86 trusts, 31 (36%) recommended OST consumption was supervised, with some instructing patients should be made to speak or swallow water to prove they were not holding medication in their mouth. Some recommended regular drug tests to monitor for illicit drug use, with one maternity guideline stating new mothers must be informed that if a test were positive, they may be discharged while their baby remains in hospital until fit for discharge. Of the 62 trusts that recommended UDS, 46 (74%) did not state a need to obtain patient consent prior to the test and six (10%) advised observing the patient urinate. Some policies advised restricting visitors, and specified patients should not be allowed to leave the ward. Four (5%) trusts recommended asking patients to sign behavioural contracts; two of which explicitly stated that medical treatment could be withdrawn if patients did not abide by contract terms.
The language used to describe patients often included potentially stigmatising terms such as ‘user’, ‘misuser’, ‘abuser’ and ‘addict’. Some policies discussed ‘sanctions’ and maintaining ‘a degree of suspicion’. Five (6%) trusts stated OST should only be initiated if patients had stopped or were motivated to change some aspects of their drug use. In some cases, policies explicitly recommended care differing from that for the general patient population, for example: ‘Patients with a history of drug abuse often have unreasonably high expectations. Alleviation of all pain is not a goal’.