Intended for healthcare professionals


Risks of interrupting drug treatment before surgery

BMJ 2000; 321 doi: (Published 23 September 2000) Cite this as: BMJ 2000;321:719

The consequences may be as serious as those from drug errors

  1. David W Noble, consultant in anaesthesia and intensive care medicine (david.noble{at},
  2. Henrik Kehlet, professor of surgery (henrik.kehlet{at}
  1. Department of Anaesthesia, Intensive Care and Hyperbaric Medicine, Grampian University Hospitals, Aberdeen AB25 2ZN
  2. Department of Surgical Gastroenterology, 435 Hvidovre University Hospital, DK-2650 Hvidovre, Denmark

    Millions of patients worldwide have surgery each year. As well as using a variety of recreational drugs, such as caffeine, tobacco, and alcohol, many of these patients are prescribed therapeutic drugs for concurrent diseases. These drugs are taken for their positive therapeutic benefits, and abrupt interruptions lead at best to a loss of effect and at worst to rebound exacerbations of diseases.1

    Unfortunately, surgery, and major abdominal surgery in particular, can cause patients to abstain from their usual drug treatment. After major abdominal surgery gastric emptying is delayed, making administration of drugs by the oral or nasogastric routes unreliable.2-4 Surgical patients may be particularly vulnerable to complications caused by abstention from drug treatment because of the adaptive and maladaptive physiological and psychological stresses related to surgical trauma that may increase the risk of organ dysfunction. 5 6 However, there is little systematic evidence that quantifies the risks of the abrupt withdrawal of therapeutic drugs in the postoperative setting.

    An observational study by Kennedy and colleagues of over 1000 admissions for general surgery and vascular surgery has made an important preliminary contribution to this field.7 It found that a high number of patients scheduled for inpatient surgery took therapeutic drugs; that there was a higher incidence of peri-operative complications among patients who took such drugs; and that there was a significant association between abstinence from therapeutic drugs and adverse outcomes.

    Half of all patients admitted to hospital were taking medicine that was unrelated to their surgery, and half of those were taking medicine for cardiovascular conditions. Only 8% were taking drugs traditionally recognised as surgically important, such as corticosteroids and drugs for diabetes. Patients taking medicine that was unrelated to their surgery were almost three times as likely to have postoperative complications as those not taking medicine, although this additional risk was smaller for drugs that were not related to cardiovascular conditions. In a logistic regression analysis there was also a highly significant association between the duration of time without medicine and complication rates. The overall complication rate for patients whose cardiovascular medicine was interrupted for less than two days was 12% compared with 27% for those whose medicine was interrupted for two days or more.

    Kennedy et al also described specific problems directly associated with the withdrawal of drugs such as antihypertensives, L-dopa, benzodiazepines, and antidepressants; these problems were successfully managed by reintroducing the therapeutic drug. These problems, which have also been documented by others, strengthen the argument for causality rather than mere association.8 Patients taking angiotensin converting enzyme inhibitors seemed to be at particular risk.

    Although reducing or abstaining from the use of recreational drugs, such as alcohol, and a small number of therapeutic drugs, such as oral anticoagulants, may be desirable before surgery, the abrupt discontinuation of most drugs is unlikely to be of any benefit. 9 10

    There are some solutions to the problems caused by abstaining from therapeutic drugs after surgery. The first solution is to allow patients to continue their usual drugs until the day of surgery when possible.11 Secondly, alternatives to the oral route of administration may be available and should be used until the oral route is re-established. Thirdly, where these alternative routes are not available for the drug then alternative drugs of the same or different class, which can be administered by a non-oral route, may be substituted. However, these solutions may not be ideal and may not be practicable. For example, in many countries angiotensin converting enzyme inhibitors are unavailable in parenteral form, and substitution of, say, another vasodilator will not produce the same effects. Fourthly, efforts to return the transit times of patients' gastrointestinal tracts to normal should be instituted as soon as possible to restore reliable drug absorption from the gut. Thus, postoperative ileus may be reduced by administering continuous epidural analgesia using local anaesthetics, avoiding unnecessary gastroduodenal tubes and restrictions on oral intake, and using non-opioid or opioid reduced analgesia together with early oral nutrition.2

    Although increased awareness and careful planning by clinicians will help reduce the adverse effects of abruptly discontinuing drugs in the postoperative period, other action is also needed. The problem of abrupt drug discontinuation is not sufficiently recognised by the pharmaceutical industry, agencies that give advice on drugs, or regulatory agencies.11 Guidance should be included in drug information leaflets and national drug formularies to help doctors identify the best substitutes for drugs that cannot be given parenterally.

    There should also be more information on the symptoms, signs, and potential sequelae of abrupt drug withdrawal and proactive strategies to help avoid these problems should be described. Pharmacists on hospital wards could play an important part in providing this guidance. Agencies that license drugs must consider requesting explicit information from manufacturers on how to manage inadvertent abstinence when considering applications for new drugs.

    There is a lack of large scale epidemiological work that precisely delineates the magnitude of peri-operative drug abstinence, but the evidence suggests that such abstinence may be harmful. More research is needed to confirm and extend the work of Kennedy et al on the impact of drug abstinence on surgical outcome and on effective methods of improving the delivery of drugs after surgery.7

    Many patients undergoing major surgery are taking therapeutic drugs. Such is the scale of the problem that failing to prevent the consequences of drug withdrawal in the postoperative period should possibly be considered equivalent to those medication errors that have been called “worse than a crime.”12


    1. 1.
    2. 2.
    3. 3.
    4. 4.
    5. 5.
    6. 6.
    7. 7.
    8. 8.
    9. 9.
    10. 10.
    11. 11.
    12. 12.
    View Abstract

    Log in

    Log in through your institution


    * For online subscription