Waiting Your Turn: Hospital Waiting Lists in Canada (10th Edition)
by Martin Zelder with Greg Wilson
Hospital waiting list survey
In order to develop a more detailed understanding of the magnitude and nature of hospital waiting lists in Canada, the authors of this study conducted a survey of specialist physicians. Specialists rather than hospital administrators were surveyed because a substantial number of hospitals either do not collect waiting list data in a systematic manner, or do not make such data publicly available (Amoko, Modrow, and Tan, 1992). In those instances where data from institutions are available, they have been used to corroborate the evidence from the survey data.
The survey was conducted in all 10 Canadian provinces. Mailing lists for the specialists polled were provided by Cornerstone List Fulfillment. The specialists on these lists are drawn from the Canadian Medical Association's membership rolls. Specialists were offered a chance to win a $2,000 prize (to be randomly awarded) as an inducement to respond. Specialists rather than general practitioners were surveyed because specialists have primary responsibility for health care management of surgical candidates. Survey questionnaires were sent to practitioners of 12 different medical specialties: plastic surgery, gynaecology, ophthalmology, otolaryngology, general surgery, neurosurgery, orthopaedic surgery, cardiovascular surgery, urology, internal medicine, radiation oncology, and medical oncology. The original survey (1990) was pretested on a sample of individual specialists serving on the relevant specialty committees of the British Columbia Medical Association. In each subsequent edition of the survey, suggestions for improvement made by responding physicians have been incorporated into the questionnaires, and in 1994, radiation oncology and medical oncology were added to the 10 specialties originally surveyed.
The questionnaire used for general surgery is found in Appendix 1. The questionnaires for all of the specialties follow this format (with a slight exception for medical oncology); only the procedures surveyed differ across the various specialty questionnaires. The data for this issue of Waiting Your Turn were collected in December 1999.
For the most part, the survey was sent to all specialists in a category. In the case of internal medicine in Ontario, approximately 500 names were randomly selected. The response rate in the five provinces initially surveyed in 1990 (British Columbia, Manitoba, New Brunswick, Newfoundland, Nova Scotia) was 20 percent. This year, the response rate was 25 percent overall, which is quite high for a mailed survey, and an increase from the 23 percent response rate of last year's survey.
The treatments identified in all of the specialist tables represent a cross-section of common procedures carried out in each specialty (definitions of procedures are found in Appendix 2). The original list of procedures was suggested by the specialty boards of the British Columbia Medical Association in 1990, and procedures have been added since then at the recommendation of survey participants.
At the suggestion of the Canadian Hospital Association, waiting time, since 1995, has been calculated as the median of physician responses rather than the mean or average, as it had been prior to 1995 (Canadian Hospital Association, 1994). The disadvantage of using average waiting times is the presence of outliers (that is, extremely long waiting times reported by a few specialists), which pull the average upwards. Changes in extreme outlier responses can have dramatic effects on the mean value even if the vast majority of the responses still cluster around the same median value. Using the median avoids this problem. The median is calculated by ranking specialists' responses in either ascending or descending order, and determining the middle value. For example, if five neurosurgeons in New Brunswick respond, the median value is the third highest (or third lowest) value among the five. 1 This means that if the median wait reported is 5 weeks for a procedure, half of the specialists reported waits of more than 5 weeks, while half of the specialists reported waits of less than 5 weeks.
The major findings from the survey responses are summarized in tables 2 through 45. Table 2 reports the median time a patient waits for an appointment with a specialist after having been referred by a general practitioner. To obtain the provincial medians (and national median) found in the last column of table 2 (and of tables 28a, 31, and 44), the 12 specialty medians are each weighted by a ratio: the number of procedures done in that specialty in the province divided by the number of total number of procedures done by specialists of all types in the province.
Tables 3 through 14 report the time a patient must wait for treatment after having seen a specialist, where the waiting time per patient is the median of the survey responses. The provincial weighted medians reported in the last line of each table are calculated by multiplying the median wait for each procedure (e.g., mammoplasty, neurolysis, etc., for plastic surgery) by a weight—the fraction of all surgeries within that specialty constituted by that procedure, with the sum of these multiplied terms forming the weighted median for that province and specialty.
Tables 15 through 26 report the estimated number of patients waiting for surgery. To allow for interprovincial comparisons, these tables also report the number of people waiting for surgery per 100,000 population. The number of people waiting for treatment is estimated using the average of the weeks waited for treatment as reported by responding specialists, and data on numbers of surgical procedures done annually from the Canadian Institute for Health Information (CIHI) for 1997-98 (Canadian Institute for Health Information, 2000; more recent versions of this report are not yet available). This report provides a count of the total number of surgical procedures performed annually in each province. To estimate the number of individuals waiting for a particular surgery, the average weeks waited for a given operation is divided by 52 and then multiplied by the total number of persons annually undergoing this particular operation. This means that a waiting period of, say, one month, implies that, on average, patients are waiting one-twelfth of a year for surgery. Therefore, the next person added to the list would find one-twelfth of a year's patients ahead of him or her in the queue. The main assumption underlying this estimate is that the number of surgeries performed will neither increase nor decrease within the year in response to waiting lists.
There are a number of minor problems in matching CIHI's categories of operations to those reported in the survey. In several instances, an operation such as rhinoplasty is listed under more than one specialty. In these cases, the number of patients annually undergoing this type of operation is divided among specialties according to the proportion of specialists in each of the overlapping specialties; e.g., if plastic surgeons constitute 75 percent of the group of specialists performing rhinoplasties, then the number of rhinoplasties counted under plastic surgery is the total multiplied by .75. A second problem is that, in some cases, an operation listed in the questionnaire has no match in the CIHI tabulation. An example is the urological operation called ureteral reimplantation for reflux, which is not listed in the CIHI count. In these cases, no estimate is made of the number of patients waiting for these operations.
Tables 28a and 28b present, respectively, median waiting time and the estimated number of patients waiting, compared among specialties and provinces. Because the questionnaires omit some procedures that are less commonly performed, the sum of the numbers of people waiting for each specialty in table 28b is, of course, an underestimate of the total number waiting. Nevertheless, the lists of procedures surveyed in the questionnaires represent between 74 percent and 83 percent of non-emergency surgery performed in each of the provinces studied.
The final row of table 28a displays the provincial and national weighted medians for the 12 specialties surveyed. As in tables 3 through 14, weighted medians are calculated by multiplying the median wait for each specialty by the fraction of all surgeries in that province occurring in that specialty, and then summing these multiplied terms corresponding to each specialty.
The number of people waiting for non-emergency surgeries that were not included in the survey was also calculated, and is listed in table 28b as the "residual" number of patients waiting. To estimate the residual number of people waiting, the number of non-emergency operations not contained in the survey that are done in each province annually must be used. This residual number of operations (compiled from the CIHI data) is then multiplied by each province's weighted average waiting time and divided by 52 (weeks). Estimates of the residual number of patients waiting are reported in table 28b.
Tables 32 through 43 report the median values for the number of weeks estimated by specialists to be clinically reasonable lengths of time to wait for treatment after an appointment with a specialist. The methodology used to construct these tables is analogous to that used in tables 3 through 14.
Verification of current data with governments
In July 2000, preliminary data were sent across Canada to provincial ministries of health, and provincial cancer and cardiac agencies. Replies were received from provincial health ministries in Alberta, British Columbia, Manitoba, New Brunswick, Prince Edward Island, and Saskatchewan, from cancer agencies in Manitoba and Ontario, and from the cardiac agency in Ontario.
The data provided allowed a number of direct comparisons to be made. In particular, 22 comparisons of waiting times were made, and 13 comparisons of numbers of people waiting were made. In 18 out of 22 waiting-time comparisons, the survey numbers were in excess of the provincially provided numbers while, in three cases, the survey numbers understated the government-estimated waiting time, and, in one case, were identical.
Of the 18 cases in which the survey data exceeded the government data, 13 were data for British Columbia. In British Columbia, the Ministry of Health defines waiting time in a manner that, by necessity, make its estimates smaller than those in this survey. Specifically, the Ministry defines a wait as the interval between the time the procedure is formally scheduled and the time it is actually carried out. Not only does this definition omit waiting time between GP and specialist (which the Institute's survey includes in the total), but it understates the patient's actual waiting time between seeing a specialist and actually receiving treatment. Nevertheless, the Ministry suggests that the degree of understatement is small: "We believe that in most cases surgeons forward... booking forms without delay once a decision to perform the procedure is taken, and that hospitals receive them within a day or two" (Kelly, 1999). However, because most hospitals only book a few months ahead, this method of measuring waiting lists undoubtedly omits a substantial fraction of patients with waits beyond the booking period (see Ramsay, 1998).
If the discrepancies between the survey and the data from British Columbia are ignored due to this difference in definitions, there are 5 remaining cases in which the survey data exceeds government estimates, three cases in which the government data exceeds the survey data, and one case in which they are identical. This approximate equality in the number of overstatements (5) and non-overstatements (3) suggests that any errors in the survey data are not substantially biased in either direction.
In addition, Saskatchewan Health (Donnelly, 2000) raises concern that the survey figures exceed their own data for most specialties. Particularly, they report that they compared average waiting times from their two major urban centres, Saskatoon and Regina, with the Institute's median waiting times, and find that for only two specialties (ophthalmology and neurosurgery) do the Institute's figures indicate a shorter waiting time. Moreover, they report that the Institute's figures are more than twice their own for gynaecology, orthopaedic surgery, otolaryngology, general surgery, and urology. They also rightly note that average waiting time will typically overstate median waiting time, implying that if they were to compute median values for their data, they would be even farther below the Institute's figures. Unfortunately, Saskatchewan Health did not provide precise numerical values, offering instead statements such as, "In gynaecology, orthopaedic surgery, otolaryngology, general surgery, and urology our information shows average waits less than half as long as the median waits you report" (Donnelly, 2000). As well, definitive comparisons are difficult to make given that the Saskatchewan Health data is urban-based, and thus not potentially representative of longer waiting times which may exist outside of urban centres (see Ramsay, 1997 for a related finding), although Saskatchewan Health offers the disclaimer that, "While smaller centres do not report waiting list information to the Department, it is our understanding that waits for surgery there are generally shorter than in the two largest districts" (Donnelly, 2000).
Finally, of the 13 cases in which numbers of patients waiting were compared, the estimates from this survey were exceeded by the government estimates in 11 cases. In other words, governments, in most cases, reckoned that more people were waiting than did participants in the Institute's survey.
Verification and comparison of earlier data with independent sources
The waiting list data can also be verified by comparison with independently-computed estimates, primarily found in academic journals. Seven studies predate the Institute's data series, and thus offer informal basis for comparison. In 1967, a survey of British Columbia hospitals was done by the British Columbia Hospital Insurance Service (Pallan, 1967). This study estimated that in 1967 the total number of people on hospital waiting lists in British Columbia exceeded 12,000—0.6 percent of the population in British Columbia that year. The 1999 estimate of 31,237 people waiting for surgery in British Columbia (a decrease of 4,148 from the 1998 estimate; see table 29b) represents 0.8 percent of the 1999 population.
A brief survey of Ontario hospitals undertaken in October 1990 for the General Accounting Office of the United States Government (1991) indicates that patients experienced waits (after seeing a specialist and before receiving treatment) for elective orthopaedic surgery ranging from 8.5 weeks to 51 weeks, for elective cardiovascular surgery ranging from one to 25 weeks, and for elective ophthalmology surgery ranging from 4.3 to 51 weeks. The new survey data presented here (in table 28a) finds typical Ontario patients waiting 13.4 weeks for orthopaedic surgery, 7.6 weeks for elective cardiovascular surgery, and 7.8 weeks for ophthalmology procedures in 1999.
A study of waiting times for radiotherapy in Ontario between 1982 and 1991 (Mackillop et al., 1994) found that the median waiting times between diagnosis by a general practitioner and initiation of radiotherapy for carcinoma of the larynx, carcinoma of the cervix, and non-small-cell lung cancer were 30.3 days, 27.2 days, and 27.3 days, respectively. The new survey data for 1999 fall within one week, roughly, of these estimates. In Ontario in 1999, the wait for radiotherapy was 42 days for each of these three cancer types (see tables 2 and 13). However, the 1999 estimate that the median wait for prostate cancer treatment was 63 days is much lower than Mackillop's estimate of 93.3 days.
A study of knee replacement surgery in Ontario found that in the late 1980s, the median wait for an initial appointment with an orthopaedic specialist was 4 weeks, while the median waiting time to receive a knee operation was 8 weeks (Coyte et al., 1994). By comparison, the Institute's survey finds that in Ontario in 1999, the wait to see an orthopaedic specialist was 8 weeks (see table 2) and the wait to receive hip or knee surgery was 16 weeks (see table 9).
Examination of waiting times for particular cardiovascular treatments in 1990 by Collins-Nakai et al. (1992) focused on three important procedures. They estimated median Canadian waiting times of 11 weeks for angioplasty and 5.5 months for cardiac bypass surgery. In comparison, 1999 median waiting times for "angiography/angioplasty" ranged from 4.5 weeks in New Brunswick to 13 weeks in Newfoundland (see table 12), and for elective cardiac bypass ranged from 8.5 weeks in Ontario to 52 weeks in Newfoundland (see table 10).
A study of waiting times for selected cardiovascular procedures in 1992 found that in Canada, 13.3 percent of waiting times for elective coronary bypass surgery fell in the 2-to-6-week range, with 40 percent in the 6-to-12-week range, 40 percent in the 12-to-24-week range, and 6.7 percent in the over-36-weeks range (Carroll et al., 1995). Again, the 1999 data indicated that the provincial waiting time for elective bypass surgery (between specialist consultation and treatment) ranged from 8.5 weeks in Ontario to 52 weeks in Newfoundland (see table 10).
Regarding waiting time for coronary artery bypass in Ontario in the early 1990s, Morgan et al. (1998) discovered that the median and mean waits were 18 and 38 days, respectively. By comparison, the 1999 Ontario survey data reveal waiting times for emergent, urgent, and elective bypass surgery of 0, 1, and 8.5 weeks, respectively (see table 10).
Three more recent studies permit direct comparison of Fraser Institute waiting times and independently-derived estimates. DeCoster et al. (1999) obtained median waiting times for 8 common surgical procedures in Manitoba for the period 1995-96. Seven of those 8 procedures—cholecystectomy, hernia repair, excision of breast lesions, varicose veins stripping and ligation, transurethral resection of the prostate, tonsillectomy, and carotid endarterectomy—are also contained in The Fraser Institute's annual survey. For 5 of these 7, Fraser Institute estimates of waiting time in Manitoba for 1995 (see Ramsay and Walker, 1996) were lower than the values found by DeCoster et al. In only one case—carotid endarterectomy—was the Institute measure higher, and for cholecystectomy the two estimates were equal.
Bell et al. (1998) surveyed the two largest hospitals in every Canadian city of 500,000 or more 2 in 1996-97 to learn their waiting times for 7 procedures, many of which were diagnostic. Among these, 3 were also collected by the Institute—magnetic resonance imaging, colonoscopy, and knee replacement. In all three cases, the median waiting times found by Bell et al. exceeded the Institute's Canada-wide waiting times (for these, see Ramsay and Walker, 1997).
Liu and Trope (1999) assessed the length of wait for selected ophthalmological surgeries in Ontario in late 1997. Three of these procedures are also tracked in the Institute's survey—cataract extraction, corneal transplant, and pterygium excision. In all three cases, the Institute figures (see Ramsay and Walker, 1998) were lower than the values independently derived by Liu and Trope.
In summary, 13 independent waiting time estimates exist for comparison with recent Institute figures. In 11 of 13 cases, the Institute figures lie below the comparison values, with only one instance in which the Institute value exceeds the comparison value, and one case in which they are identical. This evidence strongly suggests that the Institute's measurements are not biased upward, but, if anything, may be biased downward, understating actual waiting times.
Further confirmation of the magnitude of Canadian waiting times can be derived from 5 international comparative studies (the first 4 of which are noted above). Coyte et al. (1994) found that in the late 1980s, Canadians waited longer than Americans for orthopaedic consultation (5.4 vs. 3.2 weeks) and for surgery post-consultation (13.5 vs. 4.5 weeks). Collins-Nakai et al. (1992) discovered that in 1990, Canadians waited longer than Germans and Americans, respectively, for cardiac catheterization (2.2 months vs. 1.7 months vs. 0 months), angioplasty (11 weeks vs. 7 weeks vs. 0 weeks), and bypass surgery (5.5 months vs. 4.4 months vs. 0 months). Another study of cardiac procedures, by Carroll et al. (1995), revealed that in 1992 Canadians generally waited longer for both elective and urgent coronary artery bypass than did Americans (whether in private or public Veterans' Administration hospitals) and Swedes, and longer than Americans (in either hospital type) for either elective or urgent angiography. At the same time, Canadians had shorter waits than the British for elective and urgent bypasses and angiographies, and shorter waits than Swedes for both types of angiographies. Finally, Jackson, Doogue, and Elliott (1998) compared waiting time for coronary artery bypass between New Zealand in 1994-95 and Ontario in the same period, using data from Naylor et al. (1995). They found that the New Zealand mean and median waiting times (232 and 106 days, respectively) were longer than the Canadian mean and median (34 and 17 days, respectively).
1 For an even-numbered group of respondents, say, 4 physicians, the median is the average of the two middle values —in this example, the average of the second and third highest values.
2 Although not identified by name, this list was presumably comprised by Montreal, Toronto, Winnipeg, Calgary, Edmonton, and Vancouver.
You can contact us at the above email address for any comments or information requests. Please report any dead links or technical problems.