Acta Anaesthesiol Scand 2010; 54: 11851191 Printed in Singapore.

All rights reserved

r 2010 The Authors Journal compilation r 2010 The Acta Anaesthesiologica Scandinavica Foundation ACTA ANAESTHESIOLOGICA SCANDINAVICA

doi: 10.1111/j.1399-6576.2010.02318.x

Medication discontinuity errors in the perioperative period

J. A. R.
VAN

WAES1, J. C.
1

DE

GRAAFF1, A. C. G. EGBERTS2 and W. A.

2

VAN

KLEI1

Departments of Perioperative care and Emergency Medicine and Clinical Pharmacy, University Medical Centre Utrecht, the Netherlands

Background: Inappropriate withdrawal or continuation of medication in the perioperative period is associated with an increased risk for adverse events. To reduce this risk, it is important that patients take their regular medication as prescribed. We evaluated this treatment objective by studying the frequency and reasons for errors related to medication discontinuity in the perioperative period. Methods: Patients scheduled for non-cardiac surgery were included in this cross-sectional study. Perioperative medication intake was assessed at the holding area of the operation theatre complex and on the ward during the rst 24 h after surgery. Medication intake data were obtained from medical records and by questioning patients and compared with pre-operative instructions. Results: The study included 701 patients, of whom 485 (69%) used regular medication. Medication was incorrectly taken or discontinued before surgery in 27% of the pa-

tients. In 57% of these patients, the reason for incorrect intake was an unclear or a falsely understood instruction before surgery. Post-operative medication errors occurred in 26% of the patients. Conclusion: Medication errors occur frequently in the perioperative period, even in the era of an electronic medication le. Errors in prescription, administration and intake of medication are not easily solved because no single health care professional is responsible for adequate intake of medication in surgical patients. The anaesthesiologist should take on a more prominent role in regulating perioperative medication intake in surgical patients.
Accepted for publication 28 August 2010 r 2010 The Authors Journal compilation r 2010 The Acta Anaesthesiologica Scandinavica Foundation

caring for patients undergoing surgery are responsible for the perioperative management of a patients regular medication, and incorrect withdrawal or continuation of certain medication in the perioperative period increases the risk for adverse events. For example, discontinuation of cardiac medication may increase the risk of perioperative myocardial ischaemia and incorrect administration of antidiabetics could result in perioperative hypo- or hyperglycaemia.1,2 To reduce such risks, it is important that patients take their regular medication in the perioperative period as required, according to the underlying disease and the scheduled procedure. However, in 3060% of the surgical patients, medication is incorrectly administered during the perioperative
HYSICIANS

This study has been presented as a poster presentation at the 2009 NV Annual Meeting (Dutch Association of Anesthesiology, 22 May A 2009, Maastricht, The Netherlands) and the 2009 ASA Annual Meeting (18 October 2009, New Orleans, Louisiana, USA).

period.37 Several factors may contribute to the discontinuity of regular medications. On the day of surgery, instructions about fasting, together with instructions to continue regular medication before surgery, can be confusing both for patients and for nurses, resulting in the discontinuation of medication intake. After surgery, post-operative nausea, vomiting or the presence of a nasogastric tube may complicate medication intake. Furthermore, transfer from the post-operative anaesthesia care unit to the ward at the time of regular administration of medication, together with inadequate communication, could interfere with the normal intake of drugs.2,5,6 In order to improve the management of perioperative medication, it is essential to know the frequency and the reasons for the undesired discontinuity of medication in the perioperative period. The present study examined the extent and reasons for unanticipated and undesired discontinuity of regular medications in the perioperative

1185

J. A. R. van Waes et al.

period among non-cardiac surgical patients hospitalised for at least one night after surgery.

Methods
Patients
This cross-sectional study included elective noncardiac surgery patients who underwent surgery at the University Medical Centre Utrecht (the Netherlands) in April and May 2008. Patients were hospitalised after surgery for at least one night. Patients who had not been seen for pre-operative assessment at the outpatient clinic and patients with whom communication was complicated because of language, hearing problems or mental retardation were excluded from the study. The study protocol was approved by the local hospital ethics committee, which waived the need for informed consent, as patients were not subjected to investigational actions. Patient condentiality was guaranteed according to the Dutch law on personal data protection.

rounds by nurses on the clinical ward (with a maximum of four times a day) and documented for each patient in a medication le. Another regime applies to patients who are admitted after day case surgery with discharge the morning after surgery (i.e. often within 24 h after surgery). These patients are instructed to take their own medication with them to the hospital and are self-responsible for the intake of their own medication (as if they were at home). Medication intake in these patients is not documented in a medication le.

Study procedures
About 1 h before surgery, data on the instructions of oral medication intake and actual medication intake were recorded at the holding area of the operation theatre complex. Patients were specically asked which medication they were ordered (not) to take that morning according to the provided instructions and which medication they had actually taken. Furthermore, it was asked who had given these instructions and whether or not these instructions were clear or conicting. Data on medication that should have been taken or discontinued by the patient were collected from the preoperative assessment form. Paracetamol intake was excluded from pre-operative data collection, because this drug was administered by protocol to all patients 1 h before surgery. Post-operative data were recorded from the medication le on the clinical wards up to 24 h after surgery. It was established whether the medication prescribed in the hospital was consistent with the regular medication taken at home as documented in the pre-operative assessment le, and whether this medication was administered during the rst 24 h after surgery on the clinical ward. Subsequently, patients were interviewed regarding whether the medication prescribed and administered was actually taken during these rst 24 h. If not, the reasons for discontinuity were obtained. Patients who were admitted to the ICU after surgery were excluded from post-operative data collection because of difculties in obtaining data. Parenteral drugs, laxatives, creams and unguent, eye drops, ear drops, oral contraceptives and medication only taken on demand were not taken into account. Both pre-operative and post-operative data were collected by one of the authors (J. A. R. v. W.) and four research nurses using an electronic case report form.

Routine perioperative care

After the indication for surgery has been set by the surgeon, the patient visits the outpatient pre-operative assessment clinic approximately 3 weeks before surgery. At this clinic, the health status of the patient is evaluated by an anaesthesiologist or a specialised anaesthetic nurse, who provides verbal instructions about fasting and perioperative medication.8 Approximately 1 week before admission, the patient receives a standard letter by mail with general instructions about fasting and medication intake together with the admission date. Most patients are admitted on the day of surgery, which means that they are self-responsible to follow the provided medication and fasting instructions on the day of surgery. Upon admission, the patients medication is prescribed electronically by the clinical ward physician according to the pre-operative assessment form and updated by medical history. Additional medication, such as post-operative pain medication, low-molecular-weight heparins and antimicrobial drugs, is prescribed if indicated to every patient who is admitted to a clinical ward after surgery, including for those patients who do not have any own regular medication. After surgery, administration of medication is resumed at the recovery area of the operation theatre complex or at the clinical ward. All prescribed medication is administered to the patient in regular scheduled

1186

Perioperative medication discontinuity

Outcome
The primary outcome was the percentage of patients who incorrectly discontinued their regular medication on the morning of surgery. In this, the instructions provided by the anaesthesiologist at the pre-operative assessment clinic, which were documented in the medical record, were used as a reference standard for discontinuation. We also determined the percentage of patients for whom instructions about medication intake before surgery were not clear. Furthermore, the percentage of patients in whom medication during the rst 24 h after surgery was incorrectly prescribed by the physician on the ward, incorrectly administered by the nurse and nally incorrectly taken by the patient was determined. Finally, the reasons for incorrect medication intake, both pre-operative and post-operative, were collected.

medication in own administration (that is day case surgery patients who were discharged the morning after surgery, who were self-responsible for intake of their medication) and expressed as a relative risk (RR) with a 95% condence interval (95% CI).

Results
Of all 701 patients included in this cross-sectional study, pre-operative data were available for 579 (83%) patients. Post-operative data were available for 375 (53%) patients (Table 1). Fifty percent of the patients were seen at the pre-operative assessment clinic within the 3 weeks before surgery and 43% were admitted to the hospital on the morning of surgery. Most patients (90%) underwent general anaesthesia.

Pre-operative medication intake Analysis

Data were analysed using SPSS release 15.0 for Windows. For all outcomes, percentages of patients were determined. Pearsons w2-test was used to examine the association between outcomes and owing Of the 579 patients, 408 (71%) used regular medication, of whom 111 (27%) incorrectly took or discontinued this medication before surgery (Fig. 1). In most cases, medication was discontinued incorrectly. Figure 1 shows the percentage of patients

Table 1
Baseline characteristics. All patients (N 5 701) Male gender Mean age in years (SD) Available data Only pre-operative Only post-operative Both Median time between pre-operative assessment and surgery in days (IQR) Medication use Median number of medications (IQR) Admission On the day of surgery  1 day before surgery Anaesthetic technique General anaesthesia Spinal anaesthesia Loco regional anaesthesia Surgical specialty General Gynaecology and urology Orthopaedics Plastic Vascular Ear, nose, throat and dental Neurosurgery Ophthalmology Thoracic 330 (47) 55 (18) 326 (47) 122 (17) 253 (36) 22 (846) 485 (69) 2 (05) Pre-operative (N 5 579) 280 (48) 56 (18) Post-operative (N 5 375) 153 (41) 54 (18)

21 (743) 408 (71) 2 (06) 248 (43) 331 (57)

28 (850) 256 (68) 2 (06)

633 (90) 51 (7) 17 (2) 129 (18) 94 (13) 75 (11) 46 (7) 49 (7) 166 (24) 70 (10) 39 (6) 33 (5)

520 (90) 45 (8) 14 (2) 108 (19) 78 (13) 67 (12) 37 (6) 35 (6) 131 (23) 63 (11) 30 (5) 30 (5)

336 (90) 31 (8) 8 (2) 73 (20) 55 (15) 48 (13) 25 (7) 32 (7) 91 (24) 22 (6) 21 (6) 8 (2)

Figures are numbers of patients (%), unless otherwise specied. SD, standard deviation; IQR, inter quartile range.

1187

J. A. R. van Waes et al.

Fig. 1 Medication intake before surgery. Each column shows the percentage of patients who (in)correctly took or discontinued their regular medications, divided into six groups. The number of patients who used regular medication is given for each group of medication above each column. The number of patients of all six groups together exceeds the total number of patients on regular medications (N 5 408), because most patients used two or more medications and were therefore counted in more than one group. If a patient used more than one drug from the same group, of which one was correctly taken and one was incorrectly discontinued, (s)he was counted in incorrectly withdrawn within that group of medications.

who (in)correctly took or discontinued their regular medication, divided into six groups of medication. Analgesics and anticoagulation drugs were never taken incorrectly, but six (14%) of the 43 patients on oral antidiabetics took this medication incorrectly (Fig. 1). In 13 cases (3%), it was not possible to nd out whether medication should have been taken or not or which medication was actually taken. For 122 (30%) of all 408 patients who used regular medication, instructions about medication intake before surgery were not clear. Of the 111 patients who incorrectly took or discontinued their medication, in 65 cases (59%), the reason for incorrect intake was an unclear or a falsely understood instruction (Table 2). In none of the 111 patients who incorrectly took or stopped their medications the surgery was delayed because of medication discontinuation. Furthermore, none of the patients who incorrectly took their oral antidiabetics developed hypoglycaemia.

Post-operative medication intake

Of the 375 patients for whom post-operative data were available, 276 (74%) used regular medication (Table 3). In 270 (72%) of the 375 patients, one or more of four medication gifts were not recorded because data collection took place within 24 h after surgery (i.e. in general, data on the last gift were not recorded). In 98 (26%) patients, any medication error occurred during the rst 24 h after surgery, and in 82 (22%), an error occurred in the administration of medication (gifts and/or intake). In 174 (46%) of the 276 patients using regular medication, this medication was prescribed on the clinical ward and 102 (27%) patients managed their regular medication themselves. In 23 (13%) of the

Table 2
Reasons for the incorrect intake of medication before surgery (n 5 111). Instruction unclear or falsely understood Not administered on clinical ward Forgotten Contradictory instructions Refusal by patient Lack of time Other  1 reason Figures are numbers of patients (%). 65 17 10 6 4 4 3 2 (59) (15) (9) (5) (4) (4) (3) (2)

Table 3
Medication intake after surgery. All patients (N 5 375) Regular medication use Prescribed on clinical ward In self-management No regular medication use 276 (74) 174 (46) 102 (27) 99 (26) Incorrect prescription of regular medication* (N 5 23) 23 (8z) 23 (13z) NA NA Incorrect administration or intake of prescribed medicationw (N 5 82) 73 (26z) 40 (23z) 33 (32z) 9 (11z)

Given are the numbers of patients for whom regular medication was incorrectly prescribed on the clinical ward (middle column) and the numbers of patients in whom the prescribed medication was incorrectly administered or not taken after surgery (right column). Figures are numbers of patients (%). *Incorrect prescription means that one or more of the regular medications were not prescribed or that there was a discrepancy in drug or dose. wIncorrect administration or intake means that the prescribed medication was not administered or not taken by the patient. zFigures are row percentages.

1188

Perioperative medication discontinuity

Table 4
Medication that was incorrectly prescribed, administered or taken after surgery. Incorrect prescription of regular medication* (N 5 23) Analgesics, other than opiods b-Blocker Diuretic Calcium channel blocker ACE-inhibitor or ATII-receptorantagonist Statins Pulmonary medication Gastric medication Antiepileptics or psychopharmaca Antidiabetics Anticoagulation Other/not specied 2 1 3 0 4 1 2 2 6 0 0 2 (9) (4) (13) (17) (4) (9) (9) (26) (9) Incorrect administration or intake of prescribed medicationw (N 5 82) 38 1 2 2 2 3 1 2 6 0 0 25 (46) (1) (2) (2) (2) (4) (1) (2) (7) (30)

The left column shows the number of patients (%) in whom medication was incorrectly prescribed on the clinical ward, and which medications. The right column shows the number of patients (%) in whom medication was incorrectly administered or taken. Figures are numbers of patients (%). *Incorrect prescription means that one or more of the regular medications were not prescribed or there was a discrepancy in the drug or the dose. wIncorrect administration or intake means that the prescribed medication was not administered or not taken by the patient.

Table 5
Reasons for incorrect administration or intake of medication after surgery* (n 5 82). Unclear whether medication should be taken or not Nausea or vomiting Not received by the patient Nasogastric tube/Not being able to swallow Refusal by the patient Medication in self-management, but not available Other or not specied 13 6 6 3 3 2 49 (16) (7) (7) (4) (4) (2) (60)

Figures are numbers of patients (%). *Incorrect administration or intake means that the prescribed medication was not administered or not taken by the patient.

174 patients who had their regular medication prescribed on the ward, an error occurred in the prescription. Thirty-three patients (32%) who managed their medication themselves had errors in intake, compared with 40 (23%) of the patients who had their medication prescribed by the ward physician (RR 1.6, 95% CI: 0.92.8) (Table 3). The medication involved in errors included all kinds of medication (Table 4). The reasons for the incorrect administration or intake of medication after surgery were, among others, that it was unclear whether the regular medication should have been continued or not, or nausea and vomiting (Table 5).

Discussion
This study showed that errors in the prescription, administration or intake of medication in the peri-

operative period occur in a substantial number of patients. We found that surgery was not delayed in any of the patients who incorrectly took or stopped their medications before surgery. In this study, none of the patients incorrectly took anticoagulants while they should not have. In such cases, however, surgery may have been postponed by the surgeon before the patient was sent to the holding area of the operation theatre complex. Therefore, these patients were probably not included in our study. The incidence of medication errors in our study (27%) is lower than reported earlier (3060%),37 possibly because instructions about the continuation of medication around surgery have improved over the last decade. Previously, fasting was reported as the main cause of error in medication intake.46 In the present study, medication discontinuity was mainly (59%) caused by apparently confusing instructions for fasting and continuation of medication intake. During the rst 24 h after surgery, errors in medication prescription, administration or intake occurred in 26% of the patients. Kluger et al.5 found similar gures; one third of the surgical patients did not receive their medication on the day after surgery. Earlier studies on prescription errors reported that 1061% of the patients had at least one omission error in their prescribed medications at the time of hospital admission.9 We showed that these errors (mostly discontinuation of medication) still occur frequently, even though an electronic medication le is being used and that these errors also occur when

1189

J. A. R. van Waes et al.

the patient is self-responsible for medication management. Unfortunately, in some patients, data were collected before the last (fourth) administration of medication during the rst 24 h after surgery had taken place. In most of these cases, however, an error occurred in the rst or the second medication gift after surgery. Our post-operative data show that both physicians and nurses on the ward were unaware as to whether or not certain medication had to be continued. The post-operative oral intake of medication is also complicated by limited routes of administration in patients who cannot take medications orally after surgery, for example in maxillofacial surgery.2 This might be solved by clear instructions designed individually for each patient about continuation and alternative ways of administration of medication. Furthermore, alternative medication routes for oral medication should be mentioned and be available on the ward even before admission of the patient. We found more post-operative intake errors in patients who had their medication in self-management. A considerable number of these patients were not able to continue their regular medication as they should, because it was not clear for them whether they should continue their medication or not. Even if patients know that medication should be taken, it is reasonable that patients after such a major event as surgery forget to take their medication because of post-operative confusion by opiates and sedatives. In our opinion, these gures illustrate that post-operative patients should not manage their medication themselves. In the present study, both medical charts and answers from patients were used to obtain data about medication intake before and after surgery. Medication that was administered according to the medication le but, for whatever reason, was not taken by the patient was also reported as an error. Therefore, we could not only determine how many patients did not adequately take their regular medication but also at which point in time the error in the process of prescribing, administering and nally taking medication occurred. In previous studies, errors in this step of medication intake after surgery were not recorded. Still, obtaining data by asking the patient just before surgery might result in uncertain answers, because of stress or sedation by premedication. Because most patients are admitted to the hospital on the morning of surgery, we considered no other moment appropriate to collect these data than just before surgery. Furthermore, premedication is prescribed infre-

quently in our hospital. Asking the patient might also result in bias due to socially desirable answers, which we attempted to overcome by using interviewers who were not involved in daily care. While collecting data, in many cases (60%), it appeared to be impossible to nd out the reason for inadequate administration or intake. Surprisingly, in all these cases, it could not be obtained from nurses nor from the medication le and patients themselves were not aware why certain medication was not administered. Obviously, this nding can be considered as a result in itself: even in the era of electronic medication les, it remains difcult to document the reasons for (not) prescribing or administering regular medication to a patient. Although there might be valid reasons for doing so, other professionals involved in patient care (e.g. physicians or hospital pharmacists) will be unaware of the reasons why medication actually is (not) prescribed. The instruction unclear or falsely understood, forgotten and lack of time might be seen as patient-related causes of errors. In our opinion, however, all reported errors can be considered to be hospital-related system errors as in fact the preoperative work-up system should ensure that patients continue their medication correctly and also check and correct potential patient-related errors. Medication errors in the perioperative period are compounded by a fragmented approach to perioperative drug therapy, together with a lack of responsibility of a specic group of healthcare professionals.2 In our hospital, the ward physician, the anaesthesiologist and the surgeon all share responsibility for correct prescription of medications in the perioperative period. Preferentially, a single specialist should be responsible for the perioperative medication prescription. An interesting option to prevent errors is to involve a pharmacist in this process. Kwan et al.7 showed that the number of prescription errors related to regular medication reduced from 40% to 20% when a pharmacist took part in prescribing medication. To prevent errors in prescription, regular medications and commonly prescribed medications such as post-operative analgesics and low-molecular-weight heparins could already be prescribed when the patient is seen at the pre-anaesthesia evaluation clinic by the anaesthesiologist. At this visit, the anaesthesiologist should instruct the patient about the perioperative intake of medication, and written information should be provided. Also, because a number of patients will not fully remember these instructions by the time they have to

1190

Perioperative medication discontinuity

undergo surgery, individually targeted medication instructions should be sent with the patient by letter just before surgery. On admission, a nurse could check whether the medication is continued by the patient according to the given instructions. Medication that is incorrectly not taken could yet be administered before surgery, and information about medication that is incorrectly taken should be discussed with the anaesthesiologist. A nurse together with the patient could check whether the medication is prescribed according to the actual regular medication by using a checklist. The ward physician and a pharmacist can solve any problems with prescription. The anaesthesiologist should take on a more prominent role in regulating medication intake in the perioperative period. The visit of each patient to the pre-operative assessment clinic before surgery provides good opportunities to accomplish this.

3. Duthie DJR, Montgomery JN, Spence AA, Nimmo WS. Concurrent drug therapy in patients undergoing surgery. Anaesthesia 1987; 42: 3056. 4. Wyld R, Nimmo WS. Do patients fasting before and after operation receive their prescribed drug treatment? Br Med J 1988; 296: 744. 5. Kluger MT, Gale S, Plummer JL, Owen H. Peri-operative drug prescribing pattern and manufacturers guidelines: an audit. Anaesthesia 1991; 46: 4569. 6. Pearse R, Rajakulendran Y. Pre-operative fasting and administration of regular medications in adult patients presenting for elective surgery. Has the new evidence changed practice? Eur J Anaesthesiol 1999; 16: 5658. 7. Kwan Y, Fernandes OA, Nagge JJ, Wong GG, Huh JH, Hurn DA, Pond GR, Bajcar JM. Pharmacist medication assessments in a surgical preadmission clinic. Arch Intern Med 2007; 167: 103440. 8. van Klei WA, Hennis PJ, Moen J, Kalkman CJ, Moons KG. The accuracy of trained nurses in preoperative health assessment: results of the OPEN study. Anaesthesia 2004; 59: 9718. 9. Tam VC, Knowles SR, Cornish PL, Fine N, Marchesano R, Etchells EE. Frequency, type and clinical importance of medication history errors at admission to hospital: a systematic review. CMAJ 2005; 30: 5105.

References
1. Kennedy JM, van Rij AM, Spears GF, Pettigrew RA, Tucker IG. Polypharmacy in a general surgical unit and consequences of drug withdrawal. Br J Clin Pharmacol 2000; 49: 35362. 2. Noble DW, Webster J. Interrupting drug therapy in the perioperative period. Drug Saf 2002; 25: 48995.

Address: Judith A.R. van Waes Department of Perioperative care and Emergency Medicine University Medical Centre Utrecht Local mail Q04.2.313 PO Box 85500, 3508 GA Utrecht The Netherlands e-mail: j.a.r.vanwaes@umcutrecht.nl

1191