Cancellation of scheduled surgery is undesirable for patients and an inefficient use of resources. We prospectively collected data for 52 consecutive months in a public general hospital to estimate the prevalence and causes. The overall cancellation rate was 6.5% (2559 of 39 115 scheduled operations). Cancellation by broad category was for ‘medical reasons’ in 50%, ‘patient-related factors’ in 23%, and due to ‘administrative/logistic problems’ in 25%. The commonest specific causes within these categories were respectively: infections/fever (18%), patient did not attend (20%) and lack of theatre time (23%). This data will help direct resources to target prevention of cancellations as a result of these main problems.
Increasing patient satisfaction through efficient practice is an appropriate objective of a health care system. A high cancellation rate for elective surgical procedures makes it difficult to accomplish this . Cancellation reduces operating room efficiency and increases costs [2, 3].
However, different definitions of cancellation exist in the international literature [4–11]. Some authors define ‘cancellation’ as only those procedures that were cancelled on the day on which surgery was scheduled , whereas others also include those that were cancelled on the previous day [5, 6]. The Modernisation Agency Theatre Programme (National Health Service (NHS), UK) appears to define cancellations as those that occur after the patient has been notified of operation date . In definitions used by a number of reports [8–11], cancellations are considered to be any operation that appears in the definitive schedule list that ultimately is not performed. Some other studies group reasons for cancellations into relatively broad categories, while others simply list causes without grouping them [12, 13]. In yet other studies, the underlying decision to cancel is explored (e.g. whether anaesthetist, the surgeon, patient or hospital initiated the cancellation) [7, 8]. The Australian Council of Healthcare Standards  divides cancellations in day surgery into ‘failure to arrive’ and ‘cancellation of the procedure after arrival’. The latter is further subdivided into being due to ‘pre-existing medical conditions’, ‘acute onset medical condition’, and ‘administrative/organisational reasons’. Classifying cancellations is useful as it helps identify the weaknesses in the system that might be addressed.
Our study was designed to assess the causes for cancellation as a first step in identifying measures to reduce them. We estimated the prevalence of cancellations for elective surgical procedures, classifying causes as specific causes within defined broader categories.
This study was undertaken in a public general hospital (Feb 2002 – May 2006), which subsequently has become a university hospital. Details of the hospital are available at http://www.madrid.org/cs/Satellite?language=es&pagename=HospitalFundacionHospitalAlcorcon%2FPage%2FHALC_home, but briefly it is part of the Spanish National Health Service which provides universal health coverage, resources being provided through general taxation with services free at the point of care. Physicians are salaried employees on a fixed work schedule. Each hospital has a population assigned from its geographic setting which in this case consists of 250 000 inhabitants from the southwest area of Madrid. The hospital has 448 beds and a surgical suite of 11 operating theatres in which operations in the following specialities are performed: general surgery, orthopaedics and trauma, ophthalmology, gynaecology, otorhinolaryngology (ENT), urology, vascular surgery, dermatology, and the pain unit.
The study subjects were all undergoing inpatient and day stay elective surgery. Minor ambulatory surgery was performed outside this surgical suite and was therefore not included.
We recorded all operations that appeared in the ‘definitive list schedule’ and classified a ‘cancellation’ as any operation that subsequently did not occur. Our standard hospital practice is that once surgery is indicated, the surgeon routinely requests pre-operative tests: the patient and test results are then assessed in the Anaesthesia Clinic. If necessary, patients are referred for further tests or referred to physicians for further detailed medical assessment (patients so referred are not classed as ‘cancellations’). Otherwise, patients are directly assigned to the surgical waiting list. A few patients bypass this system since they are admitted directly to the surgical departments from the surgical clinic and scheduled for surgery, or are admitted as emergency cases via the emergency department or from other medical departments. In these cases, the presurgical anaesthetic assessment is always undertaken on the ward before the patient is assigned to the ‘definitive list schedule’.
Surgical lists are managed by ‘block scheduling’: each is assigned a fixed block of time or ‘session’ (e.g. 08:00–15:00 h) into which the operations are listed. These blocks of time or sessions are not assigned to specific surgeons or teams, but are distributed across the surgical services > 1 month in advance, according to the estimated need for surgical time. The distribution of surgical sessions across the specialities thus varies somewhat from month to month depending on the overall demand for surgery (i.e. the surgical waiting list) for each speciality. Since each operating theatre list is prepared as a ‘provisional list’ one week in advance, any cancellations that occur after this provisional list is published may be managed by scheduling another patient into the vacated slot, and this can reasonably occur any time up to 14:30 h on day preceding the list. Thus after 14:30 h the next day’s list schedules are considered ‘definitive’. Therefore, for this study a cancellation was defined as one occurring anytime after 14:30 h on the day before the patient’s surgery.
We collected data on the type of surgery, patient age, sex, type of admission (inpatient or day surgery) and the specific reason for cancellation. The majority of causes for cancellation by our classification are self-evident, but some need a little explanation. There are several hospital protocols related to anticoagulant and antiplatelet therapy so ‘alterations in coagulation status’ indicates non-compliance with these protocols that led to cancellation. Cancellations due to ‘inadequate preparation’ arose because, despite the pathway described above, the pre-operative medical preparation of the patient was judged insufficient to ensure patient safety (e.g. patients who had uncontrolled hyperthyroidism before thyroid surgery). This was differentiated from ‘lack of surgical readiness’, which meant that the surgeon decided it necessary to carry out additional pre-operative tests or interventions (e.g. the original diagnosis was judged unclear or bowel preparation for major surgery was judged inadequate). An ‘anaesthetic complication’ was one arising from the anaesthetic technique such as a haematoma from retrobulbar anaesthesia before surgery. ‘Difficult airway management’ indicated that expertise or equipment was unavailable to manage patients judged to pose difficulties with airway management. Cancellations due to a ‘lack of theatre time’ arose because the duration of procedures scheduled for surgery exceeded the available operating room time (e.g. the list was overbooked with cases or an unanticipated problem caused delay on the day of surgery). All data on detailed reasons for cancellation were provided by the senior surgeon and/or the anaesthetist and prospectively recorded in a computerised database. These reasons were independently checked by one of the authors on a monthly basis for the whole study period.
We grouped causes of cancellation into three broader categories: ‘medical causes’, ‘administrative/logistic causes’ and ‘patient-related causes’.
During the 52 months of study, 39 115 operations were scheduled in 9733 theatre sessions. There were 2559 cancellations (a rate of 6.5%). In all, 92% of patients underwent a pre-operative anaesthesia assessment. A similar number of women and men underwent surgery (51% vs 49%) and the cancellation rate by gender was similar (6% vs 7%). Cancellations were, however, more common in patients aged 0–10 years (13%, n =202), followed by those aged 21–30 years (9%, n =255). Cancellations were less frequent in older age groups (71–80 years, 5%, n =378; 61–70 years 6%, n = 438). The proportion of surgical procedures scheduled for inpatients (51%, n =20 070) was only slightly higher than for day surgery (49%, n =19 045). However, cancellations were less common for inpatients (5%, n =1059) than for day surgery (8%, n =1500).
Table 1 shows the distribution of elective surgical interventions and cancellations according to surgical service. The four specialties (general surgery, orthopaedics and trauma, ophthalmology and otorhinolaryngology) accounted for 75% of the interventions and 71% of the cancellations.
Table 1. Distribution of scheduled surgical procedures and cancellations according to the surgical service involved.
Scheduled surgical procedures
Cancelled surgical procedures
Proportion of operations performed by service (%)
Cancelled operations (n)
Cancellation rate (no. of cancellations/ no. of operations by service)
Contribution to total cancellations (no. of cancellations/ total no. of cancellations)
Intensive care unit
Table 2 indicates the main causes of cancellation. By broad category, ‘medical causes’ accounted for 50%, ‘patient-related causes’ for 23%, and ‘administrative/logistic causes’ for 25%. In order, the most frequent specific causes were: ‘lack of theatre time’ 23%, ‘patient did not attend’ 20% and ‘infection/fever’ 18% (the vast majority of these due to respiratory tract infection). Together, these three causes alone accounted for 60% of all cancellations.
Table 2. Prevalence of causes for cancellations in relation to the total numbers of cancellations. Definitions of some causes for cancellations, and the broad categories used, are presented in text.
Causes for cancellations
Cancellations as % of total cancellations
Acute change in cardiac function
Acute change in pulmonary function
Incomplete pre-operative study/preparation
Non-compliance with protocol for latex allergy
Change in diagnosis/indication or lack of surgical readiness
Surgical team unavailable
Non-compliance with fasting guidelines
Difficult airway management
Lack of co-operation with/acceptance of local anaesthesia
Patient did not attend
Refusal to undergo surgery once admitted
Lack of theatre time
Lack of equipment
Lack of informed consent
Lack of recovery room or intensive care unit beds
Figure 1 shows the most common causes of cancellation by age. In patients aged 0–10 years, respiratory tract infections (n = 130) accounted for 87% of infections/fever cancellations and 64% of all cancellations in this age group. The speciality that scheduled the largest number of surgical interventions in children aged 0–10 years was ENT (n =1131; 73% of all elective surgical procedures in that age group), followed by urology (n =235; 15%) and ophthalmology (n =113; 7%).
Figure 2 shows the main causes for cancellation according to the surgical service. Non-surgical medical causes were the most common reasons for cancellation in ENT (50%), ophthalmology (40%), urology (39%), gynaecology (38%), orthopaedics and trauma (32%), vascular surgery (32%) and general surgery (29%). In dermatology (52%) and the pain unit (36%), cancellations were most often patient-related.
The main result of our study is that lack of theatre time, patient non-attendance and infections/fever are by far the commonest causes of cancellation. We explore these in turn.
Lack of theatre time has also been reported by others as an important cause, contributing to cancellation in up to 3.2% of the scheduled operations [4, 6, 10, 11, 15]. This emphasises the importance of analysing theatre (in)efficiency, especially when due to over-booking and theatre over-utilisation to prevent cancellation . An important preventive measure would logically involve booking cases onto lists using knowledge of surgical procedure duration (e.g. as estimated by surgeons and anaesthetists, or by using data from theatre logs or from the previously published data [16, 17]). It is possible that overbooked theatre lists leading to over-running are managed differently in hospitals like ours (where staff are employed on contracts of fixed working hours), as compared with institutions where salaries are more directly related to variable hours of work [18, 19]. In the former, there may be no financial incentive for staff to stay late to undertake extra cases (and hence over-running lists are curtailed by cancellation). In the latter, ‘overtime’ payments may offer some incentive to prevent cancellation, but the overall costs to the institution may be very high if overbooking is common.
Cancellations due to patient non-attendance accounted for 20% of the cancellations. When expressed as a proportion of the cases scheduled, our rate of ∼1% is within the ranges reported elsewhere, of up to 5.5% [5, 6, 11, 12, 15, 20]: any differences may be influenced by the healthcare systems. We speculate that where care is free at point of care, non-attendance may be higher than where a fee-for-service system prevails. However, we cannot locate data to support this speculation. In our hospital, our admissions service reminds patients by telephone 10 days before their operation and again 2 days before surgery. The patient is asked to notify the hospital if they cannot attend, but clearly this is not fail-safe.
Infections (mainly upper respiratory tract) were responsible for cancellation especially in children presenting for ENT surgery. This is likely inevitable, particularly perhaps in winter months, but we do not know if it also represents over-cautious practice in a non-specialist (i.e. non-paediatric) hospital.
Although cancellation for medical reasons accounted for ∼3% of all scheduled procedures (which is within the range reported by others [4–6, 8, 10–12, 15]), it represents ∼50% of all cancellations. We were surprised by this since almost all (92%) patients underwent specialist pre-assessment. Although some authors appear to dissent , most studies confirm that pre-assessment reduces the cancellation rate [4, 8, 9, 21]. For example, Rai and Pandit  reported a halving of the cancellation rate, but theirs was a dedicated short-stay elective surgical unit, with certain constraints on the type of patient admitted. It has been independently observed that patients with a lower ASA grade suffer less cancellation . Another influential factor may be the interval between pre-assessment and surgery. In our study assessment occurred 2–3 months before surgery; some have suggested that an interval of 2–4 weeks is better [4, 8, 22]. The danger of a long interval is that the patient’s condition may deteriorate; the problem with too short an interval is that there is insufficient time to optimise any relevant medical conditions before surgery. Where it is not possible to shorten a long interval, consideration should be given to a second visit closer to the time of surgery in patients with concomitant pathology (which may be expensive) or (as suggested by Rai and Pandit ) a telephone consultation (which may not be sufficiently comprehensive).
Despite robust hospital protocols for the management of several isolated problems such as anticoagulation, latex allergy, pre-operative fasting, difficult airway management, etc, these issues collectively accounted for almost 30% of cancellations. This will need further review locally. More positively, and in contrast to the findings of several other studies [11, 12, 23], cancellations due to lack of hospital beds were rare. The reasons may include our relatively high proportion of day surgery cases, an integrated local management of medical and surgical high-dependency beds, efficient allocation of hospital beds in advance, coupled with flexible allocation of these beds according to the demands of each specialty.
Many previous studies on cancellations have used retrospective data [4, 9, 15, 24], yielding somewhat different results from prospective studies , examined only a single surgical specialty [5, 12, 24], assessed only a single admission category such as day surgery [5, 6], inpatient [8, 12] or short-stay surgery , or excluded some types of cancellation (e.g. cardiac surgery, obstetric and pediatric patients, and patients operated on in same-day surgery) . Our study overcomes many of these limitations and the large study size (nearly 40 000 elective surgical procedures) over a > 4 year period is robust. While we report on the impact of medical factors on cancellations, we did not explicitly collect data on the extent of previous co-morbidity, (e.g. prevailing ASA grade) so we cannot say whether medical factors had greater impact in patients already known to suffer medical problems versus those that were previously healthy.
Cancellations have an adverse effect on the concept of theatre ‘efficiency’. Pandit et al. have modelled these with degree of theatre utilisation to yield a measure of efficiency in percentage terms. Unfortunately we did not collect data on under- or over-utilised time to be able to perform a similar calculation for comparison. Nonetheless since ‘lack of theatre time’ contributed so much to cancellation, we can estimate that the combination of list over-running with a cancellation on the list was relatively frequent. Where these two occur together, the prevailing efficiency is of the order of ∼40–60% .
Since no system can be perfect, some cancellations are inevitable but there is no general agreement on what rate is ‘acceptable’. Indeed, this may depend on the type of institution (e.g. local hospital versus tertiary referral centre dealing with complex patients), the type of surgery (where patient co-morbidities may be associated with certain surgery types), the population served (e.g. its general level of health or fitness), and the healthcare system (e.g. public versus private, where patients make a more direct investment in their surgery). Basson et al.  noted that before their study, the Veterans Health Administration set 20% as the acceptable upper limit of cancellation. In New South Wales (Australia), the expected rate of cancellations is < 1.5% . In UK, rates of up to ∼10% seem to prevail . The lack of standardised definition and classification probably underlies the relatively wide range of rates of cancellation reported in the literature of between 1 and 24% [4–6, 8–12, 15, 20, 23–26].
Although hospitals will differ with respect to the services provided, personnel and infrastructures, administration and management, population served, and healthcare system in which the hospital functions, our study has relevance more widely. Our study should serve as emphasis that each hospital can usefully ascertain its main causes of cancellation. As we have discussed above, the three main causes of cancellation we found locally are similar to those reported elsewhere. Therefore we suggest that special attention is paid to (a) patient non-attendance, (b) early identification of and robust protocols for upper respiratory tract infections, (c) booking surgical lists rationally using the known or estimated times for the operations scheduled, to avoid over- or under-booking, (d) defining an optimum interval between pre-assessment and surgery. Finally, a standardised definition and classification of cancellations will facilitate data comparison between centres.