Cancellations of elective surgery may cause an inferior postoperative course: the ‘invisible hand’ of health-care prioritization?

Introduction

When resources in health care (money, time, surgeons, nurses, etc.) are limited, prioritizations have to be made. A common feature of the efforts to deal with this in a manner that can be considered fair is that priorities should be based on medical need. ^1,2 Fairness does also require openness and transparency with regard to the process of setting priorities. Where general principles give no guidance, one needs a fair process in order to solve disputes among the different views advocated. ³ Someone disagreeing about the best outcome or being the object of an unfavourable decision should be able to look into the reasons for not being regarded as the highest priority. That emergency cases in surgery should be given priority is uncontroversial, but ethics also requires one to assess the harm done to those who are down-prioritized and the reasons for it. Shortage of resources at a given time is not something indispensable, but a result of political, organizational and administrative decisions. Adam Smith ⁴ believed that an invisible hand created by the integrated forces of self-interest, competition and the supply and demand at hand would result in the best allocation of resources. However, if significant interests are at stake, as in health-care settings, one should be able to scrutinize how these forces make their concealed priorities. Organizational administrative decisions are not ethically neutral. They do not automatically lead to the best and most fair allocation.

One such situation may be when patients scheduled for elective surgery are cancelled due to an overwhelming number of emergency cases. Such cancellations of planned surgical procedures are well-known phenomena mostly regarded as administrative/economic nuisances. ^5–9 When surgeries are cancelled, for whatever reason, the hospital loses money, waiting lists grow and efficiency decreases.

From an ethical point of view such cancellations represent a dilemma. At the time of scheduling, an elective patient is given a precise point in time on which to focus all her psychological stamina. She is about to accept and allow a frightful, dangerous and potentially lethal procedure. When this surgery is cancelled, often on precariously short notice, there is a tremendous let-down and disappointment. ¹⁰

Nevertheless, the ethical dilemma often evolves to the disadvantage of the elective patient. It is a well known fact that patients in need of acute surgery after, for example, a femoral hip fracture have a greater risk of suffering from postoperative complications in cases where the operative procedure is unduly postponed. In hip fracture patients, after a delay of more than 24 hours, complications such as infections and decubital ulcers increase. ^11,12 An increased risk of postoperative mortality in the short term as well as within one year has also been reported. ¹³

In a probing pilot study, we wanted to investigate whether cancellation of elective surgeries could possibly result in the same medical complications and worse postoperative results. Hence we hypothesised that prior cancellation of elective surgery for administrative reasons may be followed by an inferior postoperative course when the planned surgical procedure takes place at a somewhat later point in time. Will a psychologically less than optimally prepared patient experience more complications and inferior final results? To study this we conducted a retrospective case-control comparison between patients who had undergone a total joint replacement with or without a prior postponement of surgery for administrative reasons.

Patients and methods

Study patients were selected retrospectively from planning lists for surgery at the department for the period 1 January 2005 to 31 May 2006. These dates were chosen in order to have a large enough study group and at least a six-month follow-up period after the index surgery. Inclusion criteria were a completed total hip (THR) or knee replacement(TKR) and that the original date of surgery had been cancelled for administrative reasons within 0–21 days prior to that original date. Patients cancelled for medical reasons were not included. Identification of patients fulfilling inclusion criteria was made from hospital charts and a control question, i.e. whether the cancellation was indeed for administrative reasons, was included in the subsequent questionnaire. Sixty-five patients were identified. In all cases surgery had been carried out within four weeks of the date that was originally planned but subsequently cancelled.

Another 65 patients, matched to the study patients based on age, gender, body mass index, American Society of Anaesthesiologists (ASA) classification, surgery planned and surgical date, were identified. These control patients underwent surgery on the date originally scheduled, i.e. without cancellation. Basic demographic data are shown in Table 1. The follow-up consisted of a mailing to all 130 patients in January 2007. This secured the same length of follow-up for both the study patients and the matched control patients. The envelope contained a questionnaire and the EQ-5D standardized quality-of-life questionnaire in combination with a letter explaining the purpose of the study. Both groups received exactly the same questionnaires. From 130 questionnaires sent out, 108 (83%) were returned. After telephone consultation 126 patients (97%) answered the questionnaire, with one patient excluded from the study group due to the fact that the cancellation was for medical reasons. This gave a total of 60 matched pairs with surgery of THR or TKR.

Complications enquired about, and stated in ‘lay language’, were: deep vein thrombosis, pulmonary embolism, wound infection, pneumonia, urinary tract infection, other infection, stroke, myocardial infarction, depression or other complication. The options given for answers were ‘yes’, ‘no’ or ‘cannot remember’.

Patients in the study group were asked about the information they received as to the cause of the cancellation and at what time prior to the planned surgical date this information was given: 27 (45%) patients were informed <24 hours before surgery, 8 (13%) <1 week before, 12 (20%) >1 week before, while 13 (22%) could not remember. Finally, both groups were asked to estimate the overall function of the operated joint on a 10-degree visual analogue scale (VAS) (1 = totally disabled, 10 fully functional) prior to surgery and at the time of the survey. The study was approved by the regional health authority ethical committee.

Statistical analysis

Repeated measurements analysis was used to analyse time-dependent data and statistical comparisons in order to test differences between the two groups were made by use of the Mann–Whitney U test. In order to evaluate hypotheses of variables in contingency tables, the χ ² test was used or, in the case of small expected frequencies, Fisher's exact test. In addition to that, descriptive statistics and graphical methods were used to characterize the data. The study employed multiple hypotheses testing, where each hypothesis was analysed separately and the existence of patterns in and the consistency of the results were considered in the analysis. All analyses were carried out by use of the SAS system, and the 5%, 1% and 0.1% levels of significance were considered. In the case of a statistically significant result the probability value (P value) has been given.

Results

Timing of the information that the planned surgery was cancelled varied from five minutes before surgery to several weeks. The majority (35 patients, 58%) received their information less than one week before surgery, whereas 25 were notified at least a week in advance. The relationship between the timing of the information of cancellation and complications was statistically insignificant. Fourteen of the 35 (40%) patients informed less than a week in advance had one or more complications. A total of 83 complications were reported: 50 in the study group (mean 0.83) and 33 (mean 0.55) in the control group (χ ² = 5.1, P < 0.05). When excluding depression as a complication, the numbers were 38 (mean 0.63) and 17 (mean 0.28), respectively (χ ² = 5.7, P < 0.05) (Table 2).

Of the individual complications, depression, urinary tract infection, wound infection and myocardial infarction were the most frequent. Myocardial infarction was the only single complication to reach a borderline statistical significance between the groups (χ ² = 4.1, P < 0.05, Fisher's exact P = 0.06) (Table 3).

The number of complications was not correlated to the timing of information (NS) but among the 17 patients who had two or more complications, including depression, nine (53%) received their information less than one week prior to surgery. The inclusion of patients who did not have proper information or inadequate recall yielded 15 patients in the group of multiple complications.

The difference, in favour of the control group, regarding quality of life at six months after the operation (EQ-5D) did not quite reach statistical significance (Mann–Whitney U test, P = 0,06).

Discussion

The results in this retrospective study support our hypothesis. We found that surgical cancellations resulted in significantly more complications and also a strong tendency towards inferior long-term wellbeing. The complications were of different kinds; the one most significantly differing between the groups was myocardial infarction. We suggest that there may actually be a cause–effect relationship, and in such a case common and everyday cancellations of elective surgeries in favour of emergency cases constitutes a significant ethical problem.

We were surprised to find myocardial infarction to be the one single complication that statistically differed between the groups. Patients were matched according to their ASA status. We specifically scrutinized the records of the four pairs of patients involved in this comparison. We could confirm the complication and we could find no increased medical disease burden for the patients suffering from this complication in the study group.

Overall, we note that the number of complications may appear large in this study and we point out that these are self-reported results. Patients may define complications in a less stringent way than medical professionals. We decided to perform the analysis both including and excluding ‘depression’ on the grounds that self-reporting of this condition may be particularly ‘weak’. In the final analysis, however, the ‘case-control’ situation remains: the study group reported more complications as well as a tendency towards inferior long-term results than did the controls.

Cancellation of planned surgery has received considerable attention for administrative reasons. Irrespective of cause, cancellation means lost operative resources, loss of efficiency and money. Most authors consider the problem in pure economic and administrative terms, solvable by redesigning work processes and training of staff. ^14,15 Others report problems on the part of the patients, i.e. a medical problem that postpones surgery. This, in turn, may be addressed by more elaborate medical clearance routines. ^16,17

A small number of articles have addressed the problem of cancellation of surgery from the patient's point of view. Ivarsson and co-workers have reported that patients scheduled for heart surgery react negatively on cancellation with anxiety, disappointment and fear, especially if the reason for cancellation was organizational in origin. The authors suggested a number of initiatives to ameliorate the situation of additional and prolonged waiting time, but they did not suggest that there were inferior results once surgery was performed. ^10,18

Sjöling et al. ¹⁹ studied the impact of preoperative information on anxiety level and postoperative pain after routine knee arthroplasty. They reported a correlation, implying that adequate preoperative information results in a lower state of anxiety and less postoperative pain. Further, it has been shown that both trait and state anxiety levels correlate to postoperative pain in women undergoing hysterectomy. ²⁰ In a comprehensive review, Vaughn et al. ²¹ found correlations between preoperative anxiety levels and postoperative pain. Research has shown clear correlations between anxiety and physiological processes such as lowering of immune response and delayed would healing ²² as well as the release of epinephrine causing blood vessel constriction, increased heart rate and increased blood pressure. ²³ Therefore, at least on a hypothetical and somewhat speculative basis, one can conceive of a sequence of processes starting with cancellation, anxiety and fear (of new cancellations?) to physiological effects that could explain a number of the complications reported in our study. The timing of the cancellation relative to the planned date is probably of significance. We were unable to prove this statistically, possibly due to loss of data. Almost half of the cases that were cancelled within the last week reported at least one complication and we suggest a week's advanced notice of cancellation to be a reasonable threshold. However, cancellations because of emergency cases are often done on a day-to-day basis.

According to clinical experience, administrative cancellations make patients disappointed and angry. ¹⁰ In a number of cases, the cancellations had caused prolonged sick-leave, changes in postoperative appointments with supporting persons on short notice and expenditures on alternative treatments such as physiotherapists and chiropractics. In contrast to when the cancellation is caused by some medical situation on the part of the patient, administrative cancellations make the patient feel a victim and out-of-control. We suggest that the psychological let-down from a considerably high level of preparation in this situation may be particularly detrimental. Anger because of unprofessional attitudes at the time of cancellation, insufficient information and empathy further aggravate the situation. ^18,24

This study has considerable limitations. The sample size is small. Again, the results in the study were self-reported by the patients and we could not safeguard against the possibility that the cancelled patients over-reported (perhaps from the disappointment that we link to possible physiological processes). Lay language diagnosis probably lacks precision. Further, and most importantly, the retrospective nature of the study resulted in the lack of some data due to forgetfulness and less than optimal data collection without a study protocol. Nevertheless, to our knowledge, this is the first time that the consequences of administrative cancellations have been analysed in terms of purely medical complications and postoperative results. Compared with a matched group of non-cancelled patients, those who had their original surgical procedure cancelled for non-medical reasons appeared to fare worse once the operative procedure had taken place.

That acute care should be prioritized is the understanding of all ethical guidelines for setting priorities when resources are limited. Prioritizations are also made between different diagnoses in accordance with medical need or expected quality of life. The common feature of these prioritizations is that they are made openly as a result of open political discussions with transparent guiding principles. Everyday administrative decisions in hospital care are typically not evaluated from an ethical point of view. Our results suggest that these decisions should not be taken lightly. Vital interests, e.g. risk of severe complications, anxiety and psychological distress, are at stake when administrative decisions to postpone treatment are taken. Respect for autonomy with a recognition of a patient's wish to be in control is also a vital ethical interest to take into consideration. We are not arguing that the priorities should change in favour of those on the list for elective surgery, but one should be aware of the ethical cost of administrative decisions of this kind. There are consequences of administrative decisions that should be openly discussed and new ways may be needed to make these priorities.

In conclusion, the results of this retrospective study represent a first support to our hypothesis that surgical cancellation is medically detrimental. Should these findings hold true in larger prospective studies, the ethical dilemma of emergency versus elective surgery becomes considerably more acute and the need to evaluate the consequences of this administrative concealed prioritization becomes urgent from an ethical point of view.