The US Food and Drug Administration has reported a tripling of drug shortages—from 61 drugs in 2005 to 178 drugs in 2010.a Among other broadly used medications, a substantial shortage of neostigmine was experienced in the United States in the last quarter of 2011.1,2 The shortage triggered media attention directed at understanding the etiology and attempting to find solutions. In an editorial appearing in Anesthesia & Analgesia, Drs. Dutton and Cohen1 wrote: “Medication shortages make us unwitting contestants in a bad episode of ‘Iron Chef,’ challenged to produce a tasty dish (a safe patient outcome) using a limited set of tools and ingredients … When should the unavailability of a desired medication lead us to postpone or cancel a scheduled case? … When the risk is significant, of course, we should include a discussion of these issues in our informed consent.” For ethical guidance, Dr. Jacobs offered, “ … if the anesthesiologist judges the risk of increased morbidity or mortality by using alternative medications to be negligible, then there is no need to discuss this issue when obtaining informed consent. However, if the anesthesiologist judges the added risk to be significant … discussion of alternative plans should be part of the informed consent process.”b Public disclosures of information regarding medical product shortages are also encouraged, and the Food and Drug Administration has specific recommendations regarding medical product shortages and guidelines for public disclosure of such information.c However, the proposed solutions do not address the fact that the decision, as suggested earlier, is made by the clinician on behalf of the patient rather than by the patient based on his or her own opinion and preference.
The shared decision-making model has been replacing other models as the desired approach to making medical care choices.3 A previous survey study regarding full disclosure in informed consent regarding postoperative visual loss (POVL) revealed that patients felt strongly about being informed about even rare (but potentially devastating) adverse events.4 The Anesthesia Patient Safety Foundation commissioned an online resource video on the issue of POVL and patient informed consent, an example of how the results of a survey study can be applied to clinical practice.d
Measuring the desire of when “reasonable patients”5 want to be informed of drug shortages and the impact of a drug shortage on clinical decisions can guide perioperative care. The goal of our study was to learn the incidence of patients wanting to be informed of drug shortages.
Mayo Clinic Arizona and Florida Sites
The eligible study population was the 949 Mayo Clinic patients who had previously undergone elective cholecystectomy surgery at either Arizona (from January 2011 to February 2014) or Florida (from January 2011 to January 2014) campus. Patients were selected from 2 Mayo locations databases to fit the criteria of (1) having undergone laparoscopic cholecystectomy as a scheduled (elective) surgery and (2) being ≥18 years old at time of the surgery. We chose the selection criteria with the intent to screen out patients with acute cholecystitis in whom the secondary questions of postponing surgery could not apply.
Once the IRB and research ethics board approvals were obtained, patients were contacted through postal mail. Included in the survey package was a cover letter explaining the purpose of the research and the study questionnaire (Supplemental Digital Content 1, Table 1, http://links.lww.com/AA/B160) to be completed and returned in a prepaid, self-addressed envelope. No incentives to participate were offered to the patients. Completion and return of the survey served as consent to participate in this study.
The shortage of neostigmine was the factor motivating our study. Therefore, we chose a surgical procedure that had a high incidence of use of neuromuscular-blocking agents (NMBAs). National US epidemiologic data regarding the use of NMBAs and their antagonists were available from Zhang et al.6 Both abdominal (80.3%) and general thoracic surgery (84.4%) had >80% use of NMBAs, in contrast with urologic, vascular, and neurologic surgery in which the rate of NMBA use was lower. On the basis of a previous study that showed demographic variables influence preference for shared medical decision making, we aimed to recruit a surgical population that had a wide range of demographic variables.7 From the Agency for Healthcare Research and Quality Clinical Classification Software (CCS),e the only 2 CCS choices that were sufficiently common were general thoracic surgery (CCS 36, lobectomy or pneumonectomy) and abdominal surgery (CCS 84, cholecystectomy and common duct exploration). The CCS 36 for “lobectomy or pneumonectomy” included multiple Current Procedural Terminology codes, including code 32657, “thoracoscopic wedge resection.” The possibility that a patient who underwent a surgical procedure (e.g., lung wedge resection) might require subsequent reoperations led to our concern that such a patient might misinterpret the study survey, both reducing survey validity and raising ethical concerns by increasing the patient’s perceived risk because of the drug shortage. We could not anticipate the results of this study until its completion and, consequently, could only poorly predict potential risks to patients (e.g., medical consequences of postponing surgery because of misunderstanding of a question). In contrast, a patient can only undergo cholecystectomy once, and, thus, we chose patients undergoing elective cholecystectomy as the study population.
The initial (pilot) survey was constructed based on face-to-face interviews with 10 patients scheduled to undergo intra-abdominal elective surgery. After conducting the pilot survey at Mayo Clinic in Florida, we refined the wording of several items in our questionnaire based on the patient feedback. The final version of the survey that was administered to patients who had previously undergone cholecystectomy was approved by the IRB at Mayo Clinic.
The primary endpoint of the survey was the percentage of patients who preferred to be given the information on drug unavailability and thus have the option to postpone elective surgery for a “slight” difference in side effects of the substituted drugs (i.e., neostigmine/glycopyrrolate substituted by edrophonium/atropine combination, necessitated by the national shortage of neostigmine). The survey’s 4-point scale included answers “definitely yes” and “probably yes.” These were combined into the binary “yes.” The study was designed to compare >50% of patients with a 60% proportion of patients who would wish to be informed of the slight differences in the drugs’ side effects. With α = 0.05, 90% statistical power would be achieved with 263 respondents. In our previous survey of Mayo Clinic patients after surgery, the response rate was 42%.4 That survey asked specific questions about informed consent. The present survey was longer and posed hypothetical questions. Estimating a response rate 1/3rd less than before, the number of patients to recruit to survey was approximately 940, where 940 = 263/(2/3 × 42%). Survey packets were mailed to patients’ listed addresses; 502 from Mayo Clinic, Florida, were mailed on May 12, 2014, and 447 were mailed from Mayo Clinic, Arizona, on June 10, 2014. The study ended on July 10, 2014.
The binomial confidence intervals (CIs) were calculated using the Clopper-Pearson method.8,9 Comparison of the observed percentage of patients wanting information with “most” (50%) patients was performed using the binomial test. Testing of the observed percentage of patients wanting information with ≥50% of patients was done based on the standard of informed consent (in some jurisdictions) being that of the expectations of a “reasonable patient.”5 By definition, at least half of patients must be “reasonable.” Thus, knowing that >50% of (reasonable) patients want to be informed of a difference in medication side effect profile would be clinically relevant. We designed our study around testing 50%, because if “most” patients want to know, that incorporates every possible choice consistent with individualized and patient-centered decision making (i.e., our study was deliberately biased against a positive result).
The survey was designed to address a range of potential responses, because we did not know what the results would be and could survey the finite population only once. We included questions for the patients’ preference “to be informed” when the side effects represented “very slight,” “slight,” or “major” differences between neostigmine/glycopyrrolate (the unavailable) combination and edrophonium/atropine (the substituting) combination. In addition, we asked the patients’ preferences to delay their surgery in case of drug shortages when the side-effect differences were “very slight,” “slight,” or “major” between the 2 drug combinations. Both increasing frequency of “yes” to wishing to be informed and “yes” to opting to postpone surgery were tested using the Cochran-Armitage trend test (the survey questionnaire is available as a secondary digital content). Because the survey was written at a relatively high reading level, evaluating the influence of educational level on the results also provided the opportunity to assess whether that may have introduced bias.
Although the responses may be valid for the population studied, that population may not reflect the rest of the US population. Three demographic variables for medical conditions were found previously to be significantly (P < 0.01) predictive of “patients’ preferences for shared decision making.”7 These were sex, age, and educational level. The Wilcoxon-Mann-Whitney U test was used to compare each of these 3 variables between the 2 groups of the primary endpoint: “yes,” wanting to be told of a drug shortage, versus “no,” not wanting to be told. We used age and educational categories in the survey and worded male and female as being “sex” rather than “gender” to match precisely the information available from the US Census Bureau. Our plan, as described in the IRB application, was to determine which of the sex, age, and educational category variables predicted the dependent variable. We would then use that knowledge to extrapolate from our sample an estimate of the nationwide percentage of patients undergoing cholecystectomy who want to be informed of drug shortage and of expected greater side effects associated with drug substitutions.
Windsor Regional Hospital Ouellette Campus, Windsor, Ontario, Canada
The Mayo Clinic has a system of same-day appointments, in which surgery is often scheduled the following day. In contrast, in Canada, the median number of days from receipt of the surgical consult to surgical visit is approximately 22 days; elective laparoscopic surgery is then performed a median of 50 days later.10 After approval of the Research Ethics Board at Windsor Regional Hospital, Windsor, Ontario, the survey was sent on May 8, 2014, to 630 patients at Windsor Regional Hospital, Ouellette Campus, to evaluate the generalizability of findings to a population of patients with likely longer surgical wait times than patients in the United States. The Canadian mail package included a consent form to be signed by the patient. The Fisher exact test was used to compare the primary endpoint (“yes” versus “no” in preference to be provided the information) between US and Canadian patients.
The statistical analyses were performed using StatXact-10 (Cytel Inc., Cambridge, MA). All P values were calculated using Monte Carlo simulation with accuracy within 0.0001. Surveys were entered into the Research Electronic Data Capture electronic data capture tool hosted at Mayo Clinic.
The response rate was 27.2% (n = 258/949). From Mayo Clinic, Florida, among the 502 surveys mailed, 163 were returned, but 13 were undeliverable and 5 subjects declined to participate. From Mayo Clinic, Arizona, among the 447 surveys mailed, 133 were returned, but 15 were undeliverable and 5 subjects declined to participate.
Most (>50%) of the respondents answering the primary question wanted “to be told by the anesthesia doctor about (the) shortage” “if there were slight differences (in side effects) between the drug combinations” (P < 0.0001, n = 256). The percentage of patients wanting to know was 76.2% (95% CI, 70.5%–81.3%).
Secondary findings suggest the reliability of our primary result. More than half of the patients (“definitely yes”) wanted to be told about the shortage if there were slight differences in side effects of the substituted drugs (60.9%, P = 0.0006). In addition, more than half (73.2%) wanted to be told if there were “very slight” differences between drugs, described to patients as being analogous to the difference between acetaminophen and aspirin for treating a headache (P = 0.0001, n = 246). Finally, very few of the patients who reported wanting to be told about the shortage if there were “very slight” differences in side effects of the substituted drugs responded that they would not want to be told if there were “slight” differences (0.8%, n = 245).
Secondary findings suggest the validity of our primary result. There was a progressive increase in the patients’ desire to gain information as the relative severity of potential side effects because of drug substitution increased (P < 0.0001; 73.2%, 76.2%, and 95.7% of 246, 256, and 253 respondents, respectively). There was a progressive increase in the patients’ preference to delay surgery (P < 0.0001; 33.6%, 39.4%, and 80.9% of 238, 246, and 241 respondents, respectively), even for a period of 1 month, as the severity of potential side effects because of drug substitution increased (P < 0.0001; 8.0%, 11.4%, and 36.1%, respectively). As the severity of the potential side-effect differences increased, patients wanting to know of the shortage but who remained willing to continue with scheduled surgery decreased: “very slight” difference, 39.6% (95% CI, 31.1%–47.5%); “slight” difference, 36.7% (95% CI, 28.4%–44.6%); and “major” difference, 14.7% (95% CI, 9.2%–20.6%).
Secondary findings suggest the generalizability of our findings to other populations of patients undergoing laparoscopic cholecystectomy. There was no significant association of patients’ responses with patients’ sex (P = 0.19), age (P = 0.76), educational level (P = 0.39), or country (United States versus Canada; P = 0.87, n = 58).f The full data are posted online as Supplemental Digital Content 1, Table 2, http://links.lww.com/AA/B160).
A recent review article reiterated the required components of informed consent for anesthesia and highlighted the areas of weakness in current practice.11 The most commonly applied standard on informational elements (i.e., disclosure and understanding) is that of the “reasonable person,” protected through statute and case law in Canada and the United States.11 Our results show that more than half of patients want to be informed of a neostigmine drug shortage when there are “slight” differences between the scarce drug and its replacement. Even if the differences in side-effect profile were “very slight,” like the difference between acetaminophen and aspirin taken for a headache, more than half the patients want to be informed (P < 0.0001).
Readers particularly focused on the importance of patient-centered decision making may consider our finding of “most” patients to be unnecessarily broad, because decision making should be individualized. In other words, many, and perhaps most, readers would consider shared decision making to be relevant, even if we had found that 10%, 25%, or 35% of patients wanted to be told about a shortage. We tested >50% to be conservative statistically, because the finding of >50% directly tested for all possible reasonable choices, including 10%, 25%, and 35%. Having tested against 50% is important, because obtaining information may represent a low cost to the patient, but not to society. Yet, 39.4% of respondents would consider delaying their surgery (i.e., a decision that would influence the patient and the cost of health care). Thus, if the standard applied was not that of a reasonable person (≥50%) but one of patient-centered decision making (e.g., 10%, 25%, or 35%), our conclusions would be stronger, not less strong.
Our results are likely generalizable findings based not only on their lack of sensitivity to age, sex, education, and country but because they match the conclusions from other studies. Corda et al.4 surveyed US patients who had spine surgery and found that at least 80% wanted to know the risk of the rare event of POVL. Hwang et al.12 surveyed patients in Korea scheduled to undergo elective outpatient orthopedic surgery. The anesthetic choices were ultrasound-guided regional anesthesia, general anesthesia, and (for lower extremity procedures) spinal anesthesia. At least half the patients wanted the choice of the anesthetic method to be a collaborative decision, defined as one for which the patient discussed the options with the anesthesiologist and in which the patient and anesthesiologist made the decision together. Asehnoune et al.13 showed that patients want information of all types: 88% of patients want information on postoperative pain, 77% on different types of anesthesia, and 63% of patients want information on all possible complications of anesthesia.
We asked patients whether they would consider delaying their scheduled surgical date without having information regarding the potential risks of postponement. The cancellation of surgeries by patients who opt to wait for an end of a drug shortage may have an economic impact on operating room budgets, but this has yet to be specifically measured. Perhaps, in future investigations, the idea of providing economic cost data to the decisions of multiple patients delaying surgery for similar periods could be studied.14–16 The University of Utah pediatric hospital told parents that the open technique for appendectomy carried lower operative material charges than laparoscopic technique and found that parents then chose more appendectomies for their children by the open surgical technique. In addition, 31% of the patients’ parents stated that hospital charges were their primary reason for choosing the open surgical technique over the laparoscopic technique.17
One of the limitations of the present survey is the method of presenting risk information to patients. Some methods (e.g., icon arrays or pictographs) influence both patients’ risk perception and risk recall18 and may help patients make more informed decisions about their health care. However, the aim of our survey was to explore whether patients wished to know about drug shortages, not to establish the optimal way in which the risk information should be presented to them. As described in the Methods, we had no previous knowledge of the percentages to be obtained. Knowing our results post hoc, future exploratory surveys and prospective studies based on the International Patient Decision Aids Standards are needed to clarify the impact that various decision aids may have on shared decision making.
Another limitation is that we have not specifically sought to establish what changes to the current patient informed consent process, if any, are needed. The current survey investigated patients’ desires to know of drug shortages, and whether such knowledge might be factored into their decisions to undergo surgery. Finally, we do not know whether the responses might have been different if the questions had been asked of people who were not scheduled to have surgery. The differences in preferences, if any, might be interesting to investigate in the future.
In conclusion, the paternalistic approach to medical decision making for patients is becoming less common. Modern medicine encourages active participation of patients and their families in their own health care decisions, and it is incumbent on physicians to understand the extent to which patients want to be informed and involved in their own health care decisions. Our results show that most patients want to be informed of drug shortages that might affect their surgical care, even if the severity of potential side effects is as small as the difference between acetaminophen and aspirin prescribed for headache.
Name: Ivan Kai-Hsiang Hsia, MD.
Contribution: This author is the survey study lead (Canada site), helped collect the data, and prepare the manuscript.
Attestation: Ivan Hsia approved the final manuscript, attests to the integrity of the original data and the analysis reported in this manuscript, and is the archival author who is responsible for maintaining the Canadian study records.
Name: Franklin Dexter, MD, PhD.
Contribution: This author helped design the study, analyze the data, and prepare the manuscript.
Attestation: Franklin Dexter attests to the integrity of the analysis reported in this manuscript and approved the final manuscript.
Name: Ilana Logvinov, RN, MSN, CCRP.
Contribution: This author helped design the study, execution of survey study, collect data (Florida campus), and review the manuscript.
Attestation: Ilana Logvinov attests to the integrity of the original data and the analysis reported in this manuscript, approved the final manuscript, and is designated as the archival author who is responsible for maintaining Florida and Arizona study records.
Name: Nikola Tankosic, MD.
Contribution: This author helped collect data (Arizona site).
Attestation: Nikola Tankosic approved the final manuscript.
Name: Harish Ramakrishna, MD, FACC.
Contribution: This author is the survey study lead (Arizona site).
Attestation: Harish Ramakrishna approved the final manuscript.
Name: Sorin J. Brull, MD, FCARCSI (Hon).
Contribution: This author helped design the study, prepare the manuscript, collect the data, and is the survey study lead (Florida campus).
Attestation: Sorin J. Brull attests to the integrity of the original data and the analysis reported in this manuscript, and approved the final manuscript.
Dr. Franklin Dexter is the Statistical Editor and Section Editor for Economics, Education, and Policy for Anesthesia & Analgesia. Dr. Sorin J. Brull is the Section Editor for Patient Safety for Anesthesia & Analgesia. This manuscript was handled by Dr. Steven L. Shafer, Editor-in-Chief, and Drs. Dexter and Brull were not involved in any way with the editorial process or decision.
a US Food and Drug Administration. A Review of FDA’s Approach to Medical Product Shortages. Washington, DC: US Department of Health and Human Services, October 31, 2011. Available at: http://www.fda.gov/DrugShortageReport. Accessed August 4, 2014.
b Jacobs JS. Practical considerations with drug shortages. Anesthesiology News 2013;121:502–6.
c Actions and Recommendations, Longer-term Action 11. Available at: http://www.fda.gov/aboutfda/reportsmanualsforms/reports/ucm277748.htm. Accessed January 27, 2015.
d http://www.apsf.org/resources/povl-consent/. Accessed August 5, 2014.
e http://FDshort.com/CCS_categories. Accessed July 23, 2014
f For example, 55% of respondents were female; 73% were aged 55 to 84 years; and 51% had at least a bachelor’s degree.
1. Dutton RP, Cohen JA. Medication shortages: are we the Iron Chefs or our own worst enemies? Anesth Analg. 2011;113:1298–9
2. De Oliveira GS Jr, Theilken LS, McCarthy RJ. Shortage of perioperative drugs: implications for anesthesia practice and patient safety. Anesth Analg. 2011;113:1429–35
3. van den Brink-Muinen A, van Dulmen SM, de Haes HC, Visser AP, Schellevis FG, Bensing JM. Has patients’ involvement in the decision-making process changed over time? Health Expect. 2006;9:333–42
4. Corda DM, Dexter F, Pasternak JJ, Trentman TL, Nottmeier EW, Brull SJ. Patients’ perspective on full disclosure and informed consent regarding postoperative visual loss associated with spinal surgery in the prone position. Mayo Clin Proc. 2011;86:865–8
5. Struder MR. The doctrine of informed consent: protecting the patient’s right to make informed health care decisions. Mont Law Rev. 1987;48:85–100
6. Zhang B, Hepner DL, Tran MH, Friedman M, Korn JR, Menzin J. Neuromuscular blockade, reversal agent use, and operating room time: retrospective analysis of US inpatient surgeries. Curr Med Res Opin. 2009;25:943–50
7. Bishop A Patients’ Preferences for Shared Decision Making: Associations with Demographic Variables, Personality Characteristics, and Characteristics of the Health Condition. 2013 Masters Thesis, University of Canterbury pages 32 and 50. Available at: http://ir.canterbury.ac.nz/handle/10092/7679
. Accessed October 20, 2014
8. Clopper CJ, Pearson ES. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika. 1934;26:404–13
9. Hahn GJ, Meeker WQ Statistical Intervals. A Guide for Practitioners. 1991 New York Wiley:82–4, 100–5
10. Lau R, Vair BA, Porter GA. Factors influencing waiting times for elective laparoscopic cholecystectomy. Can J Surg. 2007;50:34–8
11. Tait AR, Teig MK, Voepel-Lewis T. Informed consent for anesthesia: a review of practice and strategies for optimizing the consent process. Can J Anaesth. 2014;61:832–42
12. Hwang SM, Lee JJ, Jang JS, Gim GH, Kim MC, Lim SY. Patient preference and satisfaction with their involvement in the selection of an anesthetic method for surgery. J Korean Med Sci. 2014;29:287–91
13. Asehnoune K, Albaladejo P, Smail N, Heriche C, Sitbon P, Gueneron JP, Chailloleau C, Benhamou D. Information and anesthesia: what does the patient desire? (in French). Ann Fr Aneth Reanim. 2000;19:577–81
14. Dexter F, Marcon E, Epstein RH, Ledolter J. Validation of statistical methods to compare cancellation rates on the day of surgery. Anesth Analg. 2005;101:465–73
15. Dexter F, Shi P, Epstein RH. Descriptive study of case scheduling and cancellations within 1 week of the day of surgery. Anesth Analg. 2012;115:1188–95
16. Epstein RH, Dexter F. Rescheduling of previously cancelled surgi cal cases does not increase variability in operating room workload when cases are scheduled based on maximizing efficiency of use of operating room time. Anesth Analg. 2013;117:995–1002
17. Russell KW, Rollins MD, Barnhart DC, Mone MC, Meyers RL, Skarda DE, Soukup ES, Black RE, Molitor MS, Stoddard GJ, Scaife ER. Charge awareness affects treatment choice: prospective randomized trial in pediatric appendectomy. Ann Surg. 2015;262:189–93
18. Zikmund-Fisher BJ, Witteman HO, Dickson M, Fuhrel-Forbis A, Kahn VC, Exe NL, Valerio M, Holtzman LG, Scherer LD, Fagerlin A. Blocks, ovals, or people? Icon type affects risk perceptions and recall of pictographs. Med Decis Making. 2014;34:443–53