Eur Spine J DOI 10.1007/s00586-015-3878-1
ORIGINAL ARTICLE
Medication reconciliation for patients undergoing spinal surgery Pamela Kantelhardt1 • Alf Giese1 • Sven R. Kantelhardt1
Received: 1 November 2014 / Revised: 28 January 2015 / Accepted: 11 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose In recent years, a marked increase of spinal operations prompted a debate on quality issues. Besides obvious factors, such as the surgical technique, medication safety has been identified as one of the major risk factors for patients undergoing anesthesia and surgery. While the issue has already been addressed by hospital pharmacist and anesthesiologists, the prescription of correct medication remains within the surgeons’ responsibility. We, therefore, investigated medication-related errors in spinal instrumentation patients and applied current medication reconciliation strategies. Methods We performed a data survey on all patients undergoing spinal instrumentation in 2011. Risk factors for medication safety were identified and prioritized. Specific counter-measures were introduced in 2012 and evaluated in 2013. Results 147 patients were included in the 2011 and 162 in the 2013 survey. As top five risk factors we identified the preoperative stopping of medication, recording the medication history, prescription process of postoperative analgetics and anticoagulants and the medication list at discharge. Specific counter-measures included standardization of preparations, doses and the prescription process and improving access to this information (online and via a smartphone application). In elective patients, recording the The results presented here are part of the doctoral thesis of Pamela Kantelhardt. & Sven R. Kantelhardt [email protected] 1
Department of Neurosurgery, University Medical Centre Mainz, Johannes-Gutenberg University, Langenbeckstr. 1, 55131 Mainz, Germany
medication histories was delegated to a hospital pharmacist and informative flyers and posters were used to integrate the patients themselves into the process. Counter-measures directed against the first four risk factors resulted in a significant reduction of medication errors. The last risk factor was targeted by instructing the responsible staff only, which proved to be a rather ineffective measure. Conclusions Medication safety could be significantly improved by implementation of counter-measures specific to the identified risk factors. Keywords Spinal instrumentation Medication safety Medication reconciliation Hospital pharmacist
Introduction Changing demographics, high demands for the mobility of elderly patients, improved techniques and other factors have led to a marked increase in spinal operations in recent years. This prompted a debate on quality issues in spinal surgery, which is, however, part of a general discussion on quality in healthcare. Many patients are not satisfied with their healthcare systems [1] and the establishment of a new, constructive ‘‘error culture’’ is a demand often heard [2]. Quite logically, the discussion among spinal surgeons focuses on indications and technical issues. Nevertheless, several authors have investigated other factors which may influence outcome in spinal surgery patients, such as psychological factors [3] or perioperative medication [4, 5]. Among the quality issues currently under close observation by the public is medication safety. Moore et al. [6] found that 49 % of the clinical reports he reviewed contained medication errors, which increased the risk of re-hospitalizations within 3 months by a factor of 6.2. Surprisingly,
123
Eur Spine J
Unroe et al. [7] found a significantly higher rate of medication errors in surgical units compared to conservative disciplines. Orser et al. [8] describe in their review article that medication errors remain a leading cause of adverse events among all patients undergoing anesthesia (and surgery). While anesthesiologists and hospital pharmacists have started to realize and target the problem of medication safety in surgical patients, some of the problems can be better addressed by the (surgical) department to which the patients are admitted. The departments involved have unrestricted access to all patient information and it is their responsibility to prescribe the correct medication. Especially in elective patients who do not require pre- and postoperative intensive care treatment, surgeons are often the only physicians regularly seen by the patients during the course of their hospitalization. Accordingly, Hohn et al. [9] focused their study on medication safety in vascular surgery patients. We, here, specifically investigated medication-related errors in spinal instrumentation patients and demonstrate the application of current medication reconciliation strategies in this patient group.
Patients and methods
for subgroups of patients who were affected by the corresponding risk factor (elective patients only, patients requiring ICU/IMC treatment, etc.). Risk factors were then prioritized according to the frequency of their occurence and severity of their possible effects using a modified assessment scheme initially designed for flight safety which was previously adapted for medical applications [10]. Formulation of counter-measures According to the literature and guidelines, potential counter-measures were proposed by an interdisciplinary committee consisting of hospital pharmacist, neurosurgeons and nurses. This team then evaluated the counter-measures with regard to feasibility, effectiveness, efficiency and required resources and selected the most promising countermeasures to be implemented in 2012. Evaluation of counter-measures The effectiveness of the counter-measures was assessed in a retrospective survey including all patients meeting the abovementioned inclusion criteria in 2013. Results were analyzed using the same subgroups as in 2011 and the results were compared using quantifiable quality indicators.
Patients Data analysis All patients 18 years and older, who underwent spinal instrumentation in our hospital in 2011 and 2013, were included in this analysis. Patients who underwent spinal instrumentation were identified by their respective diagnosis-related groups. The 2011 survey was used for the identification of risk factors compromising medication safety. The 2013 data were used to evaluate the effectiveness of counter-measures introduced in 2012. If patients received more than one operation, only the primary intervention was analyzed. Retrospective identification of risk factors Medication errors which occurred during the hospitalization and at the hospital/outpatient care interface were retrospectively identified in a survey of all patients who met the abovementioned criteria. The investigation was performed by a hospital pharmacist. Only obvious, objectifiable errors were included (prescription not compatible with the medication information sheet and/or patient characteristics, non-adherence to contraindications, impaired renal function or interactions with other medications, dosing errors, etc.). Errors were quantified and the underlying risk factors were identified. The evaluation was performed separately
123
Quantitative data analysis was performed using the medication error databank DokuPIkR (ADKA e.V. [10, 11]). Fishers Exact Test was performed to compare the 2011 and 2013 data. Results were considered significant when p \ 0.05. Results are presented in % with the corresponding numbers of patients in brackets. All data were analyzed using the statistic software IBMÒ SPSSÒ Statistics, Version 20. Standards, checklists and other documents were posted online in the hospitals document management system (IQ Network GmbH & Co. KG, Mellrichstadt/Germany) and a self-programmed html-browser-based application for mobile terminals called ‘‘NC-APP’’ (Fig. 1).
Results Patient population Of 149 patients undergoing spinal instrumentation in 2011, two were excluded because of age \18. During the evaluation in 2013, 163 patients received spinal instrumentation. Of these, one patient had to be excluded because of incomplete documentation.
Eur Spine J Fig. 1 The NC-APP. a a smartphone running the software, b detail of the instruments screen depicting a standard process diagram
The male/female ratio was 77/70 in 2011 and 80/82 in 2013, average age was 59.6 ± 15 years in 2011 (range 20–90 years) and 62.6 ± 15.1 years in 2013 (range 21–91). In 2011, three patients died during hospitalization. 52 % of the patients (77/147) were admitted electively. In 2013, five patients died perioperatively and 42 % (68/162) were admitted electively. The only obvious difference between the two groups (2011 and 2013) concerned the percentage of patients scheduled for intensive care (ICU) or intermediate care (IMC) units postoperatively. In 2011, this was 55 % (81/ 147) and 75 % (121/162) in 2013. This, however, reflected a more cautious patient assessment by the department of anaesthesiology (the change coincided with a change in the leadership of the neuro-anesthesia team, although official standards and politics of the department of anaesthesiology did not change during the observed period). The average hospitalization time, however, did not change between 2011 and 2013. For patients requiring postoperative ICU/IMC treatment, it was 16 days in 2011 (range 5–86 days) and 15 days in 2013 (range 6–98). For patients not requiring ICU/IMC treatment, it was 7 days in both groups. Generally, the two groups were rather similar concerning demographics, secondary illnesses and surgeries. Risk factors identified in the 2011 data review Risk factor 1: preoperative stopping of medication for elective patients In 2011, 21 % of the electively admitted patients (16/77) were either on anticoagulants/antiplatelets or oral
antidiabetics, which should be stopped preoperatively in cases scheduled for elective surgery [5]. This did not happen in 19 % (3/16) of patients receiving anticoagulants/ antiplatelets or oral antidiabetics. Consequently, compared to the median preoperative hospitalization of all elective patients, which was 1 day in 2011, elective operations had to be postponed by a cumulative total amount of 11 days. Reasons for the missed discontinuation of the concerned medication were (1) lack of knowledge of preparations containing anticoagulants/antiplatelets, (2) failure to record or interpret medication history and/or missing documentation/communication. Risk factor 2: medication history for elective patients A total of 73 medication errors were related to the incorrect recording of the medication history. In total, 36 % of the electively admitted patients (28/77) did not receive the correct medication. In 93 % of these (26/28), the reasons could be clearly identified: incomplete or missing medication history (32 %, 9/28 patients) and incomplete or missing substitution of previously taken medication by drugs available in the hospital (89 %; 17 of the remaining 19 patients). In 45 % (35/77) of the electively admitted patients, a required substitution was not done. In part, nurses compensated for missing substitutions, a task for which they were, however, neither authorized nor educated. On the physicians side, lack of knowledge concerning alternative drugs available in the hospital (aut-simile) and equivalent dosage (aut-idem), and poor motivation were identified as reasons. Additionally, we found errors (such as dosage errors, interactions, double prescriptions) in previously prescribed
123
Eur Spine J
medication, prior to admission in 12 % of the electively admitted patients (9/77), which in 2011 were never adequately corrected at admission. Risk factor 3: postoperative analgetics In 33 % of the records of patient who did not require postoperative ICU/IMC treatment (22/66), double prescriptions of analgetics were found. Only 55 % of patient records documented a prescription for analgetics (36/66). Concerning on-demand prescriptions, errors were found in 23 % (15/36) of the patient records. 6 % did not have any documented analgetic prescriptions at all (4/66). Closer analysis of the prescription process revealed that postoperative analgetics were routinely prescribed in the surgical short reports, but the surgeon had no access to the patient record when writing this report, because at this time patient records were still with the anesthesia team and the patient in the recovery room. Therefore, the prescribing surgeon lacked information concerning previously taken and/or prescribed medication and other details of the medical history. Because of this organizational deficit, many surgeons did not prescribe on-demand analgetics. This led to frequent consultations of the on-call neurosurgeon, who responded by prescribing an on-demand medication on the telephone, again without consulting the patient record. Risk factor 4: postoperative anticoagulants A total of 98 medication errors related to postoperative prescription of anticoagulants were identified in 2011. Errors mainly concerned the correct dosage, but in some cases, the prescription of anticoagulants was forgotten entirely. Analysis revealed that, paralleling the situation in analgetics, postoperative anticoagulants were prescribed on the surgical short report, without access to the patient record and the documentation of individual risk factors such as thromboembolic events, family history, etc. Risk factor 5: medication list at discharge In 2011, the hospital discharge letters frequently lacked important information: In 37 % (15/41) of the patients who preoperatively were on anticoagulants/antiplatelets and 36 % (5/14) who took oral antidiabetics (20/55), no information concerning these drugs was included in the medication list at discharge. If medication was altered during the hospitalization, in 91 %, the indication was not documented (93/102). In 25 % of the patients on antibiotics at discharge (2/8), no information regarding the intended duration of this treatment was given.
123
The underlying reason was that when composing the report, physicians routinely consulted the current medication plan only, omitting the medication history, including medications stopped prior to admission. A further problem was a lack of knowledge concerning the correct dosage and durations of antibiotics. Implementation of counter-measures Risk factor 1: preoperative stopping of medication for elective patients The first counter-measure targeted the neurosurgeons knowledge of current preparations: We redesigned a checklist for the physician, listing common preparations of active substances to be stopped preoperatively and placed lists of these preparations in each room of the outpatient clinic. Second, forgetfulness and/or lack of communication were addressed by actively involving the outpatient clinic staff and the patients. The outpatient clinic staff was advised and trained to ask for the medication history with special regard to anticoagulants/antiplatelets and oral antidiabetics when scheduling admissions on the phone. Additionally, we designed patient information flyers to be handed out to each patient at the preadmission visit in the outpatient clinic. These flyers contained information concerning which drugs have to be withdrawn prior to the elective operation. Posters containing the same information were posted in our outpatient clinic. Risk factor 2: medication history for elective patients Recording the medication history at admission (for elective patients only), reviewing of the patient records (in nonelective patients also) and twice per week medication rounds on the wards were delegated to a hospital pharmacist. Risk factor 3: postoperative analgetics Postoperative analgetics were standardized and a novel prescription sheet was designed, mandatory for all new admissions. The prescription sheet contained the standard prescriptions for basis and on-demand analgetics. Analgetics are now already prescribed at admission, when all relevant information is readily available. Standards were posted online in the hospital document management system and the NC-APP. Risk factor 4: postoperative anticoagulants As risk factors 3 and 4 are related to the same problem, both were targeted by similar counter-measures. Antithrombotics were standardized (and the standard made
Eur Spine J
available online via hospital information system and NCAPP) and the prescription process was upstreamed to the admission (using the same mandatory prescription sheet as for analgetics). Risk factor 5: medication list at discharge Neurosurgeons were sensitized to consult the original medication history when composing medication lists at discharge. Evaluation of counter-measures in the 2013 data review Risk factor 1: preoperative stopping of medication for elective patients In 2013, 21 % of the elective patients (16/68) were on anticoagulants/antiplatelets oral antidiabetics. The medication was not withdrawn in 12 % (reduction of 8 % in comparison to 2011, the difference was not statistically significant, p = 0.667). In all 2013 cases, however, there were medical reasons for not withdrawing the medication and no operation had to be postponed (0 day in comparison to 11 days in 2011). Multiplied by the daily costs for in house patients (891.55€), the 11 days lost in 2011 amount to 9807.05€, which were saved in 2013. Risk factor 2: medication history for elective patients In 2013, in 32 % of the patients (21/68), the previously taken medication was only partly or not at all substituted with the drugs listed in the hospital. This happened in 0 % (0/35) of the patients seen by the hospital pharmacist. In 2011, this happened in 45 % of the patients (35/77; no pharmacist was involved in 2011; the overall difference was not statistically significant, p = 0.123). This improvement did, however, reduce the average costs for not substituted medication per patient by 86 % (0.43€ in 2013, 2.63€ in 2011). In 82 % of the patients undergoing elective surgery (56/ 68), the previously taken medication was correctly continued (64 % in 2011, the difference was statistically significant, p = 0.015). Of patients presenting with errors in the preadmission prescriptions, these were identified in 75 % (6/8, 0 % in 2011, the difference was statistically significant, p = 0.0023). Risk factor 3: postoperative analgetics In 2013, no double prescriptions were recorded (0 of 41) in comparison to 33 % (22/66) in 2011 (the difference was statistically significant, p \ 0.0001).
100 % of the patients had prescriptions of basis and ondemand medication and all of them were correct. In 2011, 94 % had a prescription for analgetics (the difference was not statistically significant, p = 0.3), 55 % had a prescription for basic analgetics (NSAIDS; this difference was statistically significant, p \ 0.0001) and on-demand analgetics were prescribed incorrectly or not at all in 23 % of these cases (the difference was likewise statistically significant, p \ 0.0001). By standardizing, analgetics average costs could be reduced by 29 % (0.12/0.17€ per patient and day in 2013/2011). Risk factor 4: postoperative anticoagulants In 2013, 80 % of the patients who did not require ICU/IMC treatment (33/41) immediately received the optimal NHM dosage, which favorably compares to 38 % (25/66) in 2011 (the difference was statistically significant, p \ 0.0001). By standardization, costs for anticoagulants (FragminÒP and P-forte) were reduced by 27 % (0.16/0.22€ per day and patient in 2013/2011). Risk factor 5: medication list at discharge In 2013, 40 % (16/40) of the patients who preoperatively took anticoagulants/antiplatelets and 56 % who were on oral antidiabetics (9/16) had no information concerning this medication in the medication list at discharge (compared to 37 % resp. 36 % in 2011). Documentation of indications for changing the previously taken medication increased from 9 to 17 % (9/102 to 19/114) between 2011 and 2013. In 2013, in 19 % of the patients on antibiotics which had to be continued following discharge (5/26), the medication list did not contain information regarding the intended duration (25 % (6/8) in 2011). None of these differences were statistically significant. Table 1 gives an overview of the results presented here.
Discussion The present discussion on quality issues goes far beyond advice and requests made by medical societies, hospital management boards and other authorities (high 5 project of the WHO, etc.). Insurance and liability companies and last, but certainly not least, the patients themselves ask for specific measures to improve the quality of care (implementation of CIRS, M&M conferences, etc.). We decided to identify and prioritize medication-related problems through a systematic survey. The results were pooled and analyzed using a critical incident databank previously applied to monitor quality issues in a neurosurgical department [10, 11].
123
Eur Spine J Table 1 An overview of the top five prioritized risk factors, implemented counter-measures, quality indicators, results for 2011 and 2013 and the statistical significance of these findings Risk factors
Counter-measure
Quality indicator
Results 2011
Results 2013
p values
Preoperative stopping of medication
Admission checklist for physicians
Rate of elective patients under anticoagulation/antidiabetics with no stop in time
19 %
12 % (all indicated)
p = 0.667
Medication history
Postoperative analgetics
Patient flyers Posters
Medication history and medication rounds by a hospital pharmacist
Standardization of postoperative analgetics (and online availability of the standards-NC-APP)
Cumulative delay of surgery
11 days
0 days
Net costs for the delay
9807.05€
0.00€
Rate of complete substitution of prescriptions to preparations listed in the hospital for elective patients
55 %
68 %
Average costs for not substituted medication/year
2.63€/patient
0.43 %/patient
Correct continuation of previously taken medication for elective patients
64 %
82 %
p = 0.015
Rate of corrections of medication errors in the medication at admission
0%
76 %
p = 0.0023
Rate of double prescriptions of postoperative analgetics
33 %
0%
p \ 0.0001
Rate of postoperatively
Special prescription sheet
Prescriptions
94 %
100 %
p = 0.30
Upstreaming of prescription process to the admission
Prescribed basic analgetics (NSAIDS)
55 %
100 %
p \ 0.0001
Incorrect prescriptions (letter of operation) of on-demand analgetics
23 %
0%
p \ 0.0001
Postoperative anticoagulants
Same as above
Medication list at discharge
Instruction of the concerned physicians only
Average costs for analgetics/day
0.17€/patient
0.12€/patient
Rate of optimal anticoagulant prescriptions postoperatively
38 %
80 %
Average costs for anticoagulants/day
0.22€/patient
0.16€/patient
Metformin
64 %
44 %
p = 0.299
Anticoagulants/antiplatelets
63 %
60 %
p = 0.821
9%
17 %
p = 0.106
75 %
81 %
p = 1.0
p \ 0.0001
Re-start of preoperatively stopped medication at discharge
Documentation of indications for altered medication Information regarding the duration of antibiotic therapy included
Our analysis reveals some alarming facts, which, however, are in line with similar reports in the recent literature. A first risk factor was already identified at preparation of elective patients for admission in the outpatient clinic (nonelective patients cannot be prepared in advance). Van Waes et al. [12] observed problems with preoperative stopping of medication in 27 % (14 % for oral antidiabetics) and Roure Nuez et al. [13] found that physicians only correctly stopped the concerned medication in 34.2 % of the cases (in comparison to 94.9 % if medication reconciliation was performed by hospital pharmacists). Corresponding to published cost–benefit calculations [14], our data demonstrate how costly these errors can be (nearly 10.000€ additional nursing costs in 2011). Corresponding to our data, checklist [15], patient flyers and written information have
123
p = 0.123
previously been shown to be effective measures, integrating the patients themselves in the process [16, 17]. The second risk factor concerns the correct continuation of previously taken medication. 64 % of patients admitted for elective spinal instrumentation received the correct medication, which again corresponds to the literature, which reports a range of 33–73 % [18–20]. Expert opinions rate that 18–59 % of the errors are relevant [18, 20]. Meguerditchian et al. [21] found that a complete substitution of medication to preparations available in the hospital (complete medication reconciliation) is performed only 20.0–47.6 % of the time. This can be compared to the 55 % in our study. Paralleling our findings, the literature reports of a reduction of errors if hospital pharmacists take the medication history [15] and/or perform the prescription
Eur Spine J
as reported by Lovely et al. [22]. Not surprisingly, our data show that a part-time pharmacist without on-call duty cannot cover all patients. This is also the reason why we focused some of our counter-measures on elective patients only. Literature reports show that up to 85 % of elderly patients bring at least one medication error when being transferred from primary care to a hospital [23]. While we found these errors in only 5 %, the hospital pharmacist corrected 6 of 68 of these errors and thus provided an additional service to the patient. Another problem arose from the prescription process of postoperative analgetics and the availability of information concerning correct doses and prescriptions. In addition to obvious medication errors (double prescriptions), only 55 % of the patients had a prescription for a basic analgetic medication and only 77 % of the patients had a valid prescription for on-demand analgetics. Meissner et al. [24] found that only 85 % of postoperative patients had an analgetic prescription. Like already described by Lee et al. [25], correct prescriptions can help to prevent the onset of pain in patients undergoing spinal surgery. Standardization of analgetics (according to the WHO), online availability via NC-APP and hospital information system and upstreaming of the prescription process proved effective measures for not only reducing the rate of errors to 0 % but also reducing the costs. The rate of correct dosage for anticoagulants (risk factor number 4) could be increased by our intervention from 38 to 80 %. Whereas an impressive number of studies address the optimal dosage and timing of postoperative anticoagulants [4, 26], we did not find any published material on the effectivity of the prescription process for anticoagulants in postoperative patients. Akeda et al. [27] found, however, that the optimization of the perioperative management of anticoagulants and physical measures helped to reduce the rate of thromboembolic events in spinal surgery patients. When comparing our 2011 and 2013 data, we likewise observed a significant improvement which even affected the economic side of the issue. The last issue targeted by our interventions was the medication list at discharge. While our rates of errors, respectively, omissions, in the medication list are comparable to those presented in literature [7, 28], we found that instructing the responsible staff alone was a rather ineffective measure, which again corresponds to the observations discussed in literature [29]. In contrast to our unsatisfactory results, Boockvar et al. [30] addressed the problem very efficiently, while simultaneously reducing the physicians’ workload by an electronic documentation system. Looking ahead, the consistent application of IT-based solutions for collection, documentation and transfer of relevant patient data could generally help to improve medication safety. The hospitals document management
system and NC-APP presented in this manuscript stand merely at the very beginning of this. The works of Zoni et al. [31], who successfully targeted errors of omission using an electronic documentation system and Schnipper et al. [32], who demonstrated that computer assistance increased the safety of the medication reconciliation process, present further steps towards an electronic patient record. This would collect all relevant data and provide access to all medical professionals concerned at any time, much like the introduction of Picture Archiving and Communication Systems (PACS) revolutionized access to radiological imaging in most of our hospitals [33].
Conclusion In line with the reports from other disciplines, we found that medication errors represent an underestimated risk in spinal instrumentation patients. The most significant problem concerned the preadmission medication and standard prescriptions for postoperative analgetics and antithrombotics and documentation issues. By application of current medication reconciliation strategies, namely standardization of preparation, doses (and online availability of this information) and prescription processes of frequently used medication, involvement of hospital pharmacist for taking medication histories and integration of the patients (using flyers and informative posters), we could significantly reduce the rate of medication errors in our department. Acknowledgments Programming and design of the NC-APP was performed by Stefan Kindel and Thomas Bauer. Statistical support was provided by Prof. Dr. Manfred Berres of the Institute for Medical Biostatistics, Epidemiology and Informatics (IMBEI), Mainz/ Germany. Conflict of interest
None.
References 1. Hauck K, Zhao X (2011) How dangerous is a day in hospital? Med Care 12(49):1068–1074 2. Reason J (2000) Human error: models and management. West J Med 172(6):393–396 3. Havakeshian S, Mannion AF (2013) Negative beliefs and psychological disturbance in spine surgery patients: a cause or consequence of a poor treatment outcome? Eur Spine J 22(12):2827–2835. doi:10.1007/s00586-013-2822-5 4. Cox JB, Weaver KJ, Neal DW, Jacob RP, Hoh DJ (2014) Decreased incidence of venous thromboembolism after spine surgery with early multimodal prophylaxis. J Neurosurg Spine. 21(4):677–684. doi:10.3171/2014.6.SPINE13447 5. Park HJ, Kwon KY, Woo JH (2014) Comparison of blood loss according to use of aspirin in lumbar fusion patients. Eur Spine J 23(8):1777–1782. doi:10.1007/s00586-014-3294-y
123
Eur Spine J 6. Moore C, Wisnivesky J, Williams S, McGinn T (2003) Medical errors related to discontinuity of care from an inpatient to an outpatient setting. J Gen Intern Med 18(8):646–651 7. Unroe KT, Pfeiffenberger T, Riegelhaupt S, Jastrzembski J, Lokhnygina Y, Colo´n-Emeric C (2010) Inpatient medication reconciliation at admission and discharge: a retrospective cohort study of age and other risk factors for medication discrepancies. Am J Geriatr Pharmacother 8(2):115–126. doi:10.1016/j.amjo pharm.2010.04.002 8. Orser BA, Hyland S, David U, Sheppard I, Wilson CR (2013) Review article: improving drug safety for patients undergoing anesthesia and surgery. Can J Anaesth 2:127–135 9. Hohn N, Langer S, Kalder J, Jacobs MJ, Marx G, Eisbert A (2014) Optimizing the pharmacotherapy of vascular surgery patients by medication reconciliation. J Cardiovasc Surg 55(2 Suppl 1):175–181 10. Kantelhardt P, Mu¨ller M, Giese A, Rohde V, Kantelhardt SR (2011) Implementation of a critical incident reporting system in a neurosurgical department. Cent Eur Neurosurg 72(1):15–21. doi:10.1055/s-0029-1243199 11. Amann S, Kantelhardt P (2012) Medication errors and medication reconciliation from a hospital pharmacist’s perspective. Z Evid Fortbild Qual Gesundhwes 106(10):717–722. doi:10. 1016/j.zefq.2012.11.015 12. van Waes JA, de Graaff JC, Egberts AC, van Klei WA (2010) Medication discontinuity errors in the perioperative period. Acta Anaesthesiol Scand 54(10):1185–1191. doi:10.1111/j.1399-6576. 2010.02318.x 13. Roure Nuez C, Gonzalez Navarro M, Gonzalez Valdivieso J, Fuster Barrera M, Broto Sumalla A, Sarle´ Rubı´ J, Garcı´a Eroles X (2012) Effectiveness of a perioperative chronic medication reconciliation program in patients scheduled for elective surgery. Med Clin 139(15):662–667. doi:10.1016/j.medcli.2012.04.032 14. van den Bemt PM, Postma MJ, van Roon EN, Chow MC, Fijn R, Brouwers JR (2002) Cost-benefit analysis of the detection of prescribing errors by hospital pharmacy staff. Drug Saf 25(2):135–143 15. Reeder TA, Mutnick A (2008) Pharmacist- versus physicianobtained medication histories. Am J Health Syst Pharm 65(9):857–860. doi:10.2146/ajhp070292 16. Fries C (2013) Patienten-Flyer fu¨r eine sichere und erfolgreiche Arzneimitteltherapie. Krankenhauspharmazie 34(1):7–10 17. Tannenbaum C, Martin P, Tamblyn R, Benedetti A, Ahmed S (2014) Reduction of inappropriate benzodiazepine prescriptions among older adults through direct patient education: the EMPOWER cluster randomized trial. JAMA Intern Med 174(6):890–898. doi:10.1001/jamainternmed.2014.949 18. Tam VC, Knowles SR, Cornish PL, Fine N, Marchesano R, Etchells EE (2005) Frequency, type and clinical importance of medication history errors at admission to hospital: a systematic review. CMAJ 173(5):510–515 19. Shepherd G, Schwartz RB (2009) Frequency of incomplete medication histories obtained at triage. Am J Health Syst Pharm 66(1):65–69. doi:10.2146/ajhp080171
123
20. Vira T, Colquhoun M, Etchells E (2006) Reconcilable differences: correcting medication errors at hospital admission and discharge. Qual Saf Health Care 15(2):122–126 21. Meguerditchian AN, Krotneva S, Reidel K, Huang A, Tamblyn R (2013) Medication reconciliation at admission and discharge: a time and motion study. BMC Health Serv Res 21(13):485. doi:10. 1186/1472-6963-13-485 22. Lovely JK, Larson DW, Quast JM (2014) A clinical practice agreement between pharmacists and surgeons streamlines medication management. Jt Comm J Qual Patient Saf Jt Comm Resour 7:296–302 23. Midlo¨v P, Bergkvist A, Bondesson A, Eriksson T, Ho¨glund P (2005) Medication errors when transferring elderly patients between primary health care and hospital care. Pharm World Sci 27(2):116–120 24. Meissner W, Ullrich K, Zwacka S, Schreiber T, Reinhart K (2001) Quality management in postoperative pain therapy. Anaesthesist 50(9):661–670 25. Lee BH, Park JO, Suk KS, Kim TH, Lee HM, Park MS, Lee SH, Park S, Lee JY, Ko SK, Moon SH (2013) Pre-emptive and multimodal perioperative pain management may improve quality of life in patients undergoing spinal surgery. Pain Physician 3:E217–E226 26. Brambilla S, Ruosi C, La Maida GA, Caserta S (2004) Prevention of venous thromboembolism in spinal surgery. Eur Spine J 13(1):1–8 27. Akeda K, Matsunaga H, Imanishi T, Hasegawa M, Sakakibara T, Kasai Y, Sudo A (2014) Prevalence and countermeasures for venous thromboembolic diseases associated with spinal surgery: a follow-up study of an institutional protocol in 209 patients. Spine 10:791–797. doi:10.1097/BRS.0000000000000295 28. Himmel W, Tabache M, Kochen MM (1996) What happens to long-term medication when general practice patients are referred to hospital? Eur J Clin Pharmacol 50(4):253–257 29. Mohta N, Vaishnava P, Liang C, Ye K, Vitale M, Dalal A, Schnipper J (2012) The effects of a ‘discharge time-out’ on the quality of hospital discharge summaries. BMJ Qual Saf 21(10):885–890 30. Boockvar KS, Livote EE, Goldstein N, Nebeker JR, Siu A, Fried T (2010) Electronic health records and adverse drug events after patient transfer. Quality Saf Health Care 5:e16 31. Zoni AC (2012) Duran Garcia ME, Jimenez Munoz AB, Salomon Perez R, Martin P, Herranz Alonso A (2012) The impact of medication reconciliation program at admission in an internal medicine department. Eur J Intern Med 8:696–700 32. Schnipper JL, Hamann C, Ndumele CD et al (2009) Effect of an electronic medication reconciliation application and process redesign on potential adverse drug events: a cluster-randomized trial. Arch Intern Med 8:771–780 33. Napoli M, Nanni M, Cimarra S, Crisafulli L, Campioni P, Marano P (2003) Picture archiving and communication in radiology. Rays 28(1):73–81
ORIGINAL ARTICLE
Medication reconciliation for patients undergoing spinal surgery Pamela Kantelhardt1 • Alf Giese1 • Sven R. Kantelhardt1
Received: 1 November 2014 / Revised: 28 January 2015 / Accepted: 11 March 2015 Ó Springer-Verlag Berlin Heidelberg 2015
Abstract Purpose In recent years, a marked increase of spinal operations prompted a debate on quality issues. Besides obvious factors, such as the surgical technique, medication safety has been identified as one of the major risk factors for patients undergoing anesthesia and surgery. While the issue has already been addressed by hospital pharmacist and anesthesiologists, the prescription of correct medication remains within the surgeons’ responsibility. We, therefore, investigated medication-related errors in spinal instrumentation patients and applied current medication reconciliation strategies. Methods We performed a data survey on all patients undergoing spinal instrumentation in 2011. Risk factors for medication safety were identified and prioritized. Specific counter-measures were introduced in 2012 and evaluated in 2013. Results 147 patients were included in the 2011 and 162 in the 2013 survey. As top five risk factors we identified the preoperative stopping of medication, recording the medication history, prescription process of postoperative analgetics and anticoagulants and the medication list at discharge. Specific counter-measures included standardization of preparations, doses and the prescription process and improving access to this information (online and via a smartphone application). In elective patients, recording the The results presented here are part of the doctoral thesis of Pamela Kantelhardt. & Sven R. Kantelhardt [email protected] 1
Department of Neurosurgery, University Medical Centre Mainz, Johannes-Gutenberg University, Langenbeckstr. 1, 55131 Mainz, Germany
medication histories was delegated to a hospital pharmacist and informative flyers and posters were used to integrate the patients themselves into the process. Counter-measures directed against the first four risk factors resulted in a significant reduction of medication errors. The last risk factor was targeted by instructing the responsible staff only, which proved to be a rather ineffective measure. Conclusions Medication safety could be significantly improved by implementation of counter-measures specific to the identified risk factors. Keywords Spinal instrumentation Medication safety Medication reconciliation Hospital pharmacist
Introduction Changing demographics, high demands for the mobility of elderly patients, improved techniques and other factors have led to a marked increase in spinal operations in recent years. This prompted a debate on quality issues in spinal surgery, which is, however, part of a general discussion on quality in healthcare. Many patients are not satisfied with their healthcare systems [1] and the establishment of a new, constructive ‘‘error culture’’ is a demand often heard [2]. Quite logically, the discussion among spinal surgeons focuses on indications and technical issues. Nevertheless, several authors have investigated other factors which may influence outcome in spinal surgery patients, such as psychological factors [3] or perioperative medication [4, 5]. Among the quality issues currently under close observation by the public is medication safety. Moore et al. [6] found that 49 % of the clinical reports he reviewed contained medication errors, which increased the risk of re-hospitalizations within 3 months by a factor of 6.2. Surprisingly,
123
Eur Spine J
Unroe et al. [7] found a significantly higher rate of medication errors in surgical units compared to conservative disciplines. Orser et al. [8] describe in their review article that medication errors remain a leading cause of adverse events among all patients undergoing anesthesia (and surgery). While anesthesiologists and hospital pharmacists have started to realize and target the problem of medication safety in surgical patients, some of the problems can be better addressed by the (surgical) department to which the patients are admitted. The departments involved have unrestricted access to all patient information and it is their responsibility to prescribe the correct medication. Especially in elective patients who do not require pre- and postoperative intensive care treatment, surgeons are often the only physicians regularly seen by the patients during the course of their hospitalization. Accordingly, Hohn et al. [9] focused their study on medication safety in vascular surgery patients. We, here, specifically investigated medication-related errors in spinal instrumentation patients and demonstrate the application of current medication reconciliation strategies in this patient group.
Patients and methods
for subgroups of patients who were affected by the corresponding risk factor (elective patients only, patients requiring ICU/IMC treatment, etc.). Risk factors were then prioritized according to the frequency of their occurence and severity of their possible effects using a modified assessment scheme initially designed for flight safety which was previously adapted for medical applications [10]. Formulation of counter-measures According to the literature and guidelines, potential counter-measures were proposed by an interdisciplinary committee consisting of hospital pharmacist, neurosurgeons and nurses. This team then evaluated the counter-measures with regard to feasibility, effectiveness, efficiency and required resources and selected the most promising countermeasures to be implemented in 2012. Evaluation of counter-measures The effectiveness of the counter-measures was assessed in a retrospective survey including all patients meeting the abovementioned inclusion criteria in 2013. Results were analyzed using the same subgroups as in 2011 and the results were compared using quantifiable quality indicators.
Patients Data analysis All patients 18 years and older, who underwent spinal instrumentation in our hospital in 2011 and 2013, were included in this analysis. Patients who underwent spinal instrumentation were identified by their respective diagnosis-related groups. The 2011 survey was used for the identification of risk factors compromising medication safety. The 2013 data were used to evaluate the effectiveness of counter-measures introduced in 2012. If patients received more than one operation, only the primary intervention was analyzed. Retrospective identification of risk factors Medication errors which occurred during the hospitalization and at the hospital/outpatient care interface were retrospectively identified in a survey of all patients who met the abovementioned criteria. The investigation was performed by a hospital pharmacist. Only obvious, objectifiable errors were included (prescription not compatible with the medication information sheet and/or patient characteristics, non-adherence to contraindications, impaired renal function or interactions with other medications, dosing errors, etc.). Errors were quantified and the underlying risk factors were identified. The evaluation was performed separately
123
Quantitative data analysis was performed using the medication error databank DokuPIkR (ADKA e.V. [10, 11]). Fishers Exact Test was performed to compare the 2011 and 2013 data. Results were considered significant when p \ 0.05. Results are presented in % with the corresponding numbers of patients in brackets. All data were analyzed using the statistic software IBMÒ SPSSÒ Statistics, Version 20. Standards, checklists and other documents were posted online in the hospitals document management system (IQ Network GmbH & Co. KG, Mellrichstadt/Germany) and a self-programmed html-browser-based application for mobile terminals called ‘‘NC-APP’’ (Fig. 1).
Results Patient population Of 149 patients undergoing spinal instrumentation in 2011, two were excluded because of age \18. During the evaluation in 2013, 163 patients received spinal instrumentation. Of these, one patient had to be excluded because of incomplete documentation.
Eur Spine J Fig. 1 The NC-APP. a a smartphone running the software, b detail of the instruments screen depicting a standard process diagram
The male/female ratio was 77/70 in 2011 and 80/82 in 2013, average age was 59.6 ± 15 years in 2011 (range 20–90 years) and 62.6 ± 15.1 years in 2013 (range 21–91). In 2011, three patients died during hospitalization. 52 % of the patients (77/147) were admitted electively. In 2013, five patients died perioperatively and 42 % (68/162) were admitted electively. The only obvious difference between the two groups (2011 and 2013) concerned the percentage of patients scheduled for intensive care (ICU) or intermediate care (IMC) units postoperatively. In 2011, this was 55 % (81/ 147) and 75 % (121/162) in 2013. This, however, reflected a more cautious patient assessment by the department of anaesthesiology (the change coincided with a change in the leadership of the neuro-anesthesia team, although official standards and politics of the department of anaesthesiology did not change during the observed period). The average hospitalization time, however, did not change between 2011 and 2013. For patients requiring postoperative ICU/IMC treatment, it was 16 days in 2011 (range 5–86 days) and 15 days in 2013 (range 6–98). For patients not requiring ICU/IMC treatment, it was 7 days in both groups. Generally, the two groups were rather similar concerning demographics, secondary illnesses and surgeries. Risk factors identified in the 2011 data review Risk factor 1: preoperative stopping of medication for elective patients In 2011, 21 % of the electively admitted patients (16/77) were either on anticoagulants/antiplatelets or oral
antidiabetics, which should be stopped preoperatively in cases scheduled for elective surgery [5]. This did not happen in 19 % (3/16) of patients receiving anticoagulants/ antiplatelets or oral antidiabetics. Consequently, compared to the median preoperative hospitalization of all elective patients, which was 1 day in 2011, elective operations had to be postponed by a cumulative total amount of 11 days. Reasons for the missed discontinuation of the concerned medication were (1) lack of knowledge of preparations containing anticoagulants/antiplatelets, (2) failure to record or interpret medication history and/or missing documentation/communication. Risk factor 2: medication history for elective patients A total of 73 medication errors were related to the incorrect recording of the medication history. In total, 36 % of the electively admitted patients (28/77) did not receive the correct medication. In 93 % of these (26/28), the reasons could be clearly identified: incomplete or missing medication history (32 %, 9/28 patients) and incomplete or missing substitution of previously taken medication by drugs available in the hospital (89 %; 17 of the remaining 19 patients). In 45 % (35/77) of the electively admitted patients, a required substitution was not done. In part, nurses compensated for missing substitutions, a task for which they were, however, neither authorized nor educated. On the physicians side, lack of knowledge concerning alternative drugs available in the hospital (aut-simile) and equivalent dosage (aut-idem), and poor motivation were identified as reasons. Additionally, we found errors (such as dosage errors, interactions, double prescriptions) in previously prescribed
123
Eur Spine J
medication, prior to admission in 12 % of the electively admitted patients (9/77), which in 2011 were never adequately corrected at admission. Risk factor 3: postoperative analgetics In 33 % of the records of patient who did not require postoperative ICU/IMC treatment (22/66), double prescriptions of analgetics were found. Only 55 % of patient records documented a prescription for analgetics (36/66). Concerning on-demand prescriptions, errors were found in 23 % (15/36) of the patient records. 6 % did not have any documented analgetic prescriptions at all (4/66). Closer analysis of the prescription process revealed that postoperative analgetics were routinely prescribed in the surgical short reports, but the surgeon had no access to the patient record when writing this report, because at this time patient records were still with the anesthesia team and the patient in the recovery room. Therefore, the prescribing surgeon lacked information concerning previously taken and/or prescribed medication and other details of the medical history. Because of this organizational deficit, many surgeons did not prescribe on-demand analgetics. This led to frequent consultations of the on-call neurosurgeon, who responded by prescribing an on-demand medication on the telephone, again without consulting the patient record. Risk factor 4: postoperative anticoagulants A total of 98 medication errors related to postoperative prescription of anticoagulants were identified in 2011. Errors mainly concerned the correct dosage, but in some cases, the prescription of anticoagulants was forgotten entirely. Analysis revealed that, paralleling the situation in analgetics, postoperative anticoagulants were prescribed on the surgical short report, without access to the patient record and the documentation of individual risk factors such as thromboembolic events, family history, etc. Risk factor 5: medication list at discharge In 2011, the hospital discharge letters frequently lacked important information: In 37 % (15/41) of the patients who preoperatively were on anticoagulants/antiplatelets and 36 % (5/14) who took oral antidiabetics (20/55), no information concerning these drugs was included in the medication list at discharge. If medication was altered during the hospitalization, in 91 %, the indication was not documented (93/102). In 25 % of the patients on antibiotics at discharge (2/8), no information regarding the intended duration of this treatment was given.
123
The underlying reason was that when composing the report, physicians routinely consulted the current medication plan only, omitting the medication history, including medications stopped prior to admission. A further problem was a lack of knowledge concerning the correct dosage and durations of antibiotics. Implementation of counter-measures Risk factor 1: preoperative stopping of medication for elective patients The first counter-measure targeted the neurosurgeons knowledge of current preparations: We redesigned a checklist for the physician, listing common preparations of active substances to be stopped preoperatively and placed lists of these preparations in each room of the outpatient clinic. Second, forgetfulness and/or lack of communication were addressed by actively involving the outpatient clinic staff and the patients. The outpatient clinic staff was advised and trained to ask for the medication history with special regard to anticoagulants/antiplatelets and oral antidiabetics when scheduling admissions on the phone. Additionally, we designed patient information flyers to be handed out to each patient at the preadmission visit in the outpatient clinic. These flyers contained information concerning which drugs have to be withdrawn prior to the elective operation. Posters containing the same information were posted in our outpatient clinic. Risk factor 2: medication history for elective patients Recording the medication history at admission (for elective patients only), reviewing of the patient records (in nonelective patients also) and twice per week medication rounds on the wards were delegated to a hospital pharmacist. Risk factor 3: postoperative analgetics Postoperative analgetics were standardized and a novel prescription sheet was designed, mandatory for all new admissions. The prescription sheet contained the standard prescriptions for basis and on-demand analgetics. Analgetics are now already prescribed at admission, when all relevant information is readily available. Standards were posted online in the hospital document management system and the NC-APP. Risk factor 4: postoperative anticoagulants As risk factors 3 and 4 are related to the same problem, both were targeted by similar counter-measures. Antithrombotics were standardized (and the standard made
Eur Spine J
available online via hospital information system and NCAPP) and the prescription process was upstreamed to the admission (using the same mandatory prescription sheet as for analgetics). Risk factor 5: medication list at discharge Neurosurgeons were sensitized to consult the original medication history when composing medication lists at discharge. Evaluation of counter-measures in the 2013 data review Risk factor 1: preoperative stopping of medication for elective patients In 2013, 21 % of the elective patients (16/68) were on anticoagulants/antiplatelets oral antidiabetics. The medication was not withdrawn in 12 % (reduction of 8 % in comparison to 2011, the difference was not statistically significant, p = 0.667). In all 2013 cases, however, there were medical reasons for not withdrawing the medication and no operation had to be postponed (0 day in comparison to 11 days in 2011). Multiplied by the daily costs for in house patients (891.55€), the 11 days lost in 2011 amount to 9807.05€, which were saved in 2013. Risk factor 2: medication history for elective patients In 2013, in 32 % of the patients (21/68), the previously taken medication was only partly or not at all substituted with the drugs listed in the hospital. This happened in 0 % (0/35) of the patients seen by the hospital pharmacist. In 2011, this happened in 45 % of the patients (35/77; no pharmacist was involved in 2011; the overall difference was not statistically significant, p = 0.123). This improvement did, however, reduce the average costs for not substituted medication per patient by 86 % (0.43€ in 2013, 2.63€ in 2011). In 82 % of the patients undergoing elective surgery (56/ 68), the previously taken medication was correctly continued (64 % in 2011, the difference was statistically significant, p = 0.015). Of patients presenting with errors in the preadmission prescriptions, these were identified in 75 % (6/8, 0 % in 2011, the difference was statistically significant, p = 0.0023). Risk factor 3: postoperative analgetics In 2013, no double prescriptions were recorded (0 of 41) in comparison to 33 % (22/66) in 2011 (the difference was statistically significant, p \ 0.0001).
100 % of the patients had prescriptions of basis and ondemand medication and all of them were correct. In 2011, 94 % had a prescription for analgetics (the difference was not statistically significant, p = 0.3), 55 % had a prescription for basic analgetics (NSAIDS; this difference was statistically significant, p \ 0.0001) and on-demand analgetics were prescribed incorrectly or not at all in 23 % of these cases (the difference was likewise statistically significant, p \ 0.0001). By standardizing, analgetics average costs could be reduced by 29 % (0.12/0.17€ per patient and day in 2013/2011). Risk factor 4: postoperative anticoagulants In 2013, 80 % of the patients who did not require ICU/IMC treatment (33/41) immediately received the optimal NHM dosage, which favorably compares to 38 % (25/66) in 2011 (the difference was statistically significant, p \ 0.0001). By standardization, costs for anticoagulants (FragminÒP and P-forte) were reduced by 27 % (0.16/0.22€ per day and patient in 2013/2011). Risk factor 5: medication list at discharge In 2013, 40 % (16/40) of the patients who preoperatively took anticoagulants/antiplatelets and 56 % who were on oral antidiabetics (9/16) had no information concerning this medication in the medication list at discharge (compared to 37 % resp. 36 % in 2011). Documentation of indications for changing the previously taken medication increased from 9 to 17 % (9/102 to 19/114) between 2011 and 2013. In 2013, in 19 % of the patients on antibiotics which had to be continued following discharge (5/26), the medication list did not contain information regarding the intended duration (25 % (6/8) in 2011). None of these differences were statistically significant. Table 1 gives an overview of the results presented here.
Discussion The present discussion on quality issues goes far beyond advice and requests made by medical societies, hospital management boards and other authorities (high 5 project of the WHO, etc.). Insurance and liability companies and last, but certainly not least, the patients themselves ask for specific measures to improve the quality of care (implementation of CIRS, M&M conferences, etc.). We decided to identify and prioritize medication-related problems through a systematic survey. The results were pooled and analyzed using a critical incident databank previously applied to monitor quality issues in a neurosurgical department [10, 11].
123
Eur Spine J Table 1 An overview of the top five prioritized risk factors, implemented counter-measures, quality indicators, results for 2011 and 2013 and the statistical significance of these findings Risk factors
Counter-measure
Quality indicator
Results 2011
Results 2013
p values
Preoperative stopping of medication
Admission checklist for physicians
Rate of elective patients under anticoagulation/antidiabetics with no stop in time
19 %
12 % (all indicated)
p = 0.667
Medication history
Postoperative analgetics
Patient flyers Posters
Medication history and medication rounds by a hospital pharmacist
Standardization of postoperative analgetics (and online availability of the standards-NC-APP)
Cumulative delay of surgery
11 days
0 days
Net costs for the delay
9807.05€
0.00€
Rate of complete substitution of prescriptions to preparations listed in the hospital for elective patients
55 %
68 %
Average costs for not substituted medication/year
2.63€/patient
0.43 %/patient
Correct continuation of previously taken medication for elective patients
64 %
82 %
p = 0.015
Rate of corrections of medication errors in the medication at admission
0%
76 %
p = 0.0023
Rate of double prescriptions of postoperative analgetics
33 %
0%
p \ 0.0001
Rate of postoperatively
Special prescription sheet
Prescriptions
94 %
100 %
p = 0.30
Upstreaming of prescription process to the admission
Prescribed basic analgetics (NSAIDS)
55 %
100 %
p \ 0.0001
Incorrect prescriptions (letter of operation) of on-demand analgetics
23 %
0%
p \ 0.0001
Postoperative anticoagulants
Same as above
Medication list at discharge
Instruction of the concerned physicians only
Average costs for analgetics/day
0.17€/patient
0.12€/patient
Rate of optimal anticoagulant prescriptions postoperatively
38 %
80 %
Average costs for anticoagulants/day
0.22€/patient
0.16€/patient
Metformin
64 %
44 %
p = 0.299
Anticoagulants/antiplatelets
63 %
60 %
p = 0.821
9%
17 %
p = 0.106
75 %
81 %
p = 1.0
p \ 0.0001
Re-start of preoperatively stopped medication at discharge
Documentation of indications for altered medication Information regarding the duration of antibiotic therapy included
Our analysis reveals some alarming facts, which, however, are in line with similar reports in the recent literature. A first risk factor was already identified at preparation of elective patients for admission in the outpatient clinic (nonelective patients cannot be prepared in advance). Van Waes et al. [12] observed problems with preoperative stopping of medication in 27 % (14 % for oral antidiabetics) and Roure Nuez et al. [13] found that physicians only correctly stopped the concerned medication in 34.2 % of the cases (in comparison to 94.9 % if medication reconciliation was performed by hospital pharmacists). Corresponding to published cost–benefit calculations [14], our data demonstrate how costly these errors can be (nearly 10.000€ additional nursing costs in 2011). Corresponding to our data, checklist [15], patient flyers and written information have
123
p = 0.123
previously been shown to be effective measures, integrating the patients themselves in the process [16, 17]. The second risk factor concerns the correct continuation of previously taken medication. 64 % of patients admitted for elective spinal instrumentation received the correct medication, which again corresponds to the literature, which reports a range of 33–73 % [18–20]. Expert opinions rate that 18–59 % of the errors are relevant [18, 20]. Meguerditchian et al. [21] found that a complete substitution of medication to preparations available in the hospital (complete medication reconciliation) is performed only 20.0–47.6 % of the time. This can be compared to the 55 % in our study. Paralleling our findings, the literature reports of a reduction of errors if hospital pharmacists take the medication history [15] and/or perform the prescription
Eur Spine J
as reported by Lovely et al. [22]. Not surprisingly, our data show that a part-time pharmacist without on-call duty cannot cover all patients. This is also the reason why we focused some of our counter-measures on elective patients only. Literature reports show that up to 85 % of elderly patients bring at least one medication error when being transferred from primary care to a hospital [23]. While we found these errors in only 5 %, the hospital pharmacist corrected 6 of 68 of these errors and thus provided an additional service to the patient. Another problem arose from the prescription process of postoperative analgetics and the availability of information concerning correct doses and prescriptions. In addition to obvious medication errors (double prescriptions), only 55 % of the patients had a prescription for a basic analgetic medication and only 77 % of the patients had a valid prescription for on-demand analgetics. Meissner et al. [24] found that only 85 % of postoperative patients had an analgetic prescription. Like already described by Lee et al. [25], correct prescriptions can help to prevent the onset of pain in patients undergoing spinal surgery. Standardization of analgetics (according to the WHO), online availability via NC-APP and hospital information system and upstreaming of the prescription process proved effective measures for not only reducing the rate of errors to 0 % but also reducing the costs. The rate of correct dosage for anticoagulants (risk factor number 4) could be increased by our intervention from 38 to 80 %. Whereas an impressive number of studies address the optimal dosage and timing of postoperative anticoagulants [4, 26], we did not find any published material on the effectivity of the prescription process for anticoagulants in postoperative patients. Akeda et al. [27] found, however, that the optimization of the perioperative management of anticoagulants and physical measures helped to reduce the rate of thromboembolic events in spinal surgery patients. When comparing our 2011 and 2013 data, we likewise observed a significant improvement which even affected the economic side of the issue. The last issue targeted by our interventions was the medication list at discharge. While our rates of errors, respectively, omissions, in the medication list are comparable to those presented in literature [7, 28], we found that instructing the responsible staff alone was a rather ineffective measure, which again corresponds to the observations discussed in literature [29]. In contrast to our unsatisfactory results, Boockvar et al. [30] addressed the problem very efficiently, while simultaneously reducing the physicians’ workload by an electronic documentation system. Looking ahead, the consistent application of IT-based solutions for collection, documentation and transfer of relevant patient data could generally help to improve medication safety. The hospitals document management
system and NC-APP presented in this manuscript stand merely at the very beginning of this. The works of Zoni et al. [31], who successfully targeted errors of omission using an electronic documentation system and Schnipper et al. [32], who demonstrated that computer assistance increased the safety of the medication reconciliation process, present further steps towards an electronic patient record. This would collect all relevant data and provide access to all medical professionals concerned at any time, much like the introduction of Picture Archiving and Communication Systems (PACS) revolutionized access to radiological imaging in most of our hospitals [33].
Conclusion In line with the reports from other disciplines, we found that medication errors represent an underestimated risk in spinal instrumentation patients. The most significant problem concerned the preadmission medication and standard prescriptions for postoperative analgetics and antithrombotics and documentation issues. By application of current medication reconciliation strategies, namely standardization of preparation, doses (and online availability of this information) and prescription processes of frequently used medication, involvement of hospital pharmacist for taking medication histories and integration of the patients (using flyers and informative posters), we could significantly reduce the rate of medication errors in our department. Acknowledgments Programming and design of the NC-APP was performed by Stefan Kindel and Thomas Bauer. Statistical support was provided by Prof. Dr. Manfred Berres of the Institute for Medical Biostatistics, Epidemiology and Informatics (IMBEI), Mainz/ Germany. Conflict of interest
None.
References 1. Hauck K, Zhao X (2011) How dangerous is a day in hospital? Med Care 12(49):1068–1074 2. Reason J (2000) Human error: models and management. West J Med 172(6):393–396 3. Havakeshian S, Mannion AF (2013) Negative beliefs and psychological disturbance in spine surgery patients: a cause or consequence of a poor treatment outcome? Eur Spine J 22(12):2827–2835. doi:10.1007/s00586-013-2822-5 4. Cox JB, Weaver KJ, Neal DW, Jacob RP, Hoh DJ (2014) Decreased incidence of venous thromboembolism after spine surgery with early multimodal prophylaxis. J Neurosurg Spine. 21(4):677–684. doi:10.3171/2014.6.SPINE13447 5. Park HJ, Kwon KY, Woo JH (2014) Comparison of blood loss according to use of aspirin in lumbar fusion patients. Eur Spine J 23(8):1777–1782. doi:10.1007/s00586-014-3294-y
123
Eur Spine J 6. Moore C, Wisnivesky J, Williams S, McGinn T (2003) Medical errors related to discontinuity of care from an inpatient to an outpatient setting. J Gen Intern Med 18(8):646–651 7. Unroe KT, Pfeiffenberger T, Riegelhaupt S, Jastrzembski J, Lokhnygina Y, Colo´n-Emeric C (2010) Inpatient medication reconciliation at admission and discharge: a retrospective cohort study of age and other risk factors for medication discrepancies. Am J Geriatr Pharmacother 8(2):115–126. doi:10.1016/j.amjo pharm.2010.04.002 8. Orser BA, Hyland S, David U, Sheppard I, Wilson CR (2013) Review article: improving drug safety for patients undergoing anesthesia and surgery. Can J Anaesth 2:127–135 9. Hohn N, Langer S, Kalder J, Jacobs MJ, Marx G, Eisbert A (2014) Optimizing the pharmacotherapy of vascular surgery patients by medication reconciliation. J Cardiovasc Surg 55(2 Suppl 1):175–181 10. Kantelhardt P, Mu¨ller M, Giese A, Rohde V, Kantelhardt SR (2011) Implementation of a critical incident reporting system in a neurosurgical department. Cent Eur Neurosurg 72(1):15–21. doi:10.1055/s-0029-1243199 11. Amann S, Kantelhardt P (2012) Medication errors and medication reconciliation from a hospital pharmacist’s perspective. Z Evid Fortbild Qual Gesundhwes 106(10):717–722. doi:10. 1016/j.zefq.2012.11.015 12. van Waes JA, de Graaff JC, Egberts AC, van Klei WA (2010) Medication discontinuity errors in the perioperative period. Acta Anaesthesiol Scand 54(10):1185–1191. doi:10.1111/j.1399-6576. 2010.02318.x 13. Roure Nuez C, Gonzalez Navarro M, Gonzalez Valdivieso J, Fuster Barrera M, Broto Sumalla A, Sarle´ Rubı´ J, Garcı´a Eroles X (2012) Effectiveness of a perioperative chronic medication reconciliation program in patients scheduled for elective surgery. Med Clin 139(15):662–667. doi:10.1016/j.medcli.2012.04.032 14. van den Bemt PM, Postma MJ, van Roon EN, Chow MC, Fijn R, Brouwers JR (2002) Cost-benefit analysis of the detection of prescribing errors by hospital pharmacy staff. Drug Saf 25(2):135–143 15. Reeder TA, Mutnick A (2008) Pharmacist- versus physicianobtained medication histories. Am J Health Syst Pharm 65(9):857–860. doi:10.2146/ajhp070292 16. Fries C (2013) Patienten-Flyer fu¨r eine sichere und erfolgreiche Arzneimitteltherapie. Krankenhauspharmazie 34(1):7–10 17. Tannenbaum C, Martin P, Tamblyn R, Benedetti A, Ahmed S (2014) Reduction of inappropriate benzodiazepine prescriptions among older adults through direct patient education: the EMPOWER cluster randomized trial. JAMA Intern Med 174(6):890–898. doi:10.1001/jamainternmed.2014.949 18. Tam VC, Knowles SR, Cornish PL, Fine N, Marchesano R, Etchells EE (2005) Frequency, type and clinical importance of medication history errors at admission to hospital: a systematic review. CMAJ 173(5):510–515 19. Shepherd G, Schwartz RB (2009) Frequency of incomplete medication histories obtained at triage. Am J Health Syst Pharm 66(1):65–69. doi:10.2146/ajhp080171
123
20. Vira T, Colquhoun M, Etchells E (2006) Reconcilable differences: correcting medication errors at hospital admission and discharge. Qual Saf Health Care 15(2):122–126 21. Meguerditchian AN, Krotneva S, Reidel K, Huang A, Tamblyn R (2013) Medication reconciliation at admission and discharge: a time and motion study. BMC Health Serv Res 21(13):485. doi:10. 1186/1472-6963-13-485 22. Lovely JK, Larson DW, Quast JM (2014) A clinical practice agreement between pharmacists and surgeons streamlines medication management. Jt Comm J Qual Patient Saf Jt Comm Resour 7:296–302 23. Midlo¨v P, Bergkvist A, Bondesson A, Eriksson T, Ho¨glund P (2005) Medication errors when transferring elderly patients between primary health care and hospital care. Pharm World Sci 27(2):116–120 24. Meissner W, Ullrich K, Zwacka S, Schreiber T, Reinhart K (2001) Quality management in postoperative pain therapy. Anaesthesist 50(9):661–670 25. Lee BH, Park JO, Suk KS, Kim TH, Lee HM, Park MS, Lee SH, Park S, Lee JY, Ko SK, Moon SH (2013) Pre-emptive and multimodal perioperative pain management may improve quality of life in patients undergoing spinal surgery. Pain Physician 3:E217–E226 26. Brambilla S, Ruosi C, La Maida GA, Caserta S (2004) Prevention of venous thromboembolism in spinal surgery. Eur Spine J 13(1):1–8 27. Akeda K, Matsunaga H, Imanishi T, Hasegawa M, Sakakibara T, Kasai Y, Sudo A (2014) Prevalence and countermeasures for venous thromboembolic diseases associated with spinal surgery: a follow-up study of an institutional protocol in 209 patients. Spine 10:791–797. doi:10.1097/BRS.0000000000000295 28. Himmel W, Tabache M, Kochen MM (1996) What happens to long-term medication when general practice patients are referred to hospital? Eur J Clin Pharmacol 50(4):253–257 29. Mohta N, Vaishnava P, Liang C, Ye K, Vitale M, Dalal A, Schnipper J (2012) The effects of a ‘discharge time-out’ on the quality of hospital discharge summaries. BMJ Qual Saf 21(10):885–890 30. Boockvar KS, Livote EE, Goldstein N, Nebeker JR, Siu A, Fried T (2010) Electronic health records and adverse drug events after patient transfer. Quality Saf Health Care 5:e16 31. Zoni AC (2012) Duran Garcia ME, Jimenez Munoz AB, Salomon Perez R, Martin P, Herranz Alonso A (2012) The impact of medication reconciliation program at admission in an internal medicine department. Eur J Intern Med 8:696–700 32. Schnipper JL, Hamann C, Ndumele CD et al (2009) Effect of an electronic medication reconciliation application and process redesign on potential adverse drug events: a cluster-randomized trial. Arch Intern Med 8:771–780 33. Napoli M, Nanni M, Cimarra S, Crisafulli L, Campioni P, Marano P (2003) Picture archiving and communication in radiology. Rays 28(1):73–81
