OUP user menu

Timely access and quality of care in colorectal cancer: are they related?

Yarrow J. McConnell, Karen Inglis, Geoff A. Porter
DOI: http://dx.doi.org/10.1093/intqhc/mzq010 219-228 First published online: 5 March 2010

Abstract

Objective Colorectal cancer patients want both timely access and high-quality care. The objective of this study was to explore relationships between quality indicators and access time intervals specific to colorectal cancer patients.

Design Prospective consecutive cohort study.

Setting Single health district.

Participants Between February 2002 and February 2004, all patients undergoing non-emergent surgery for primary colorectal cancer were enrolled.

Intervention A standardized method was used to collect clinicodemographic, diagnostic and treatment event data.

Main Outcome Measures Associations between accepted colorectal cancer-specific quality indicators and benchmarked access time intervals for diagnosis, surgery and adjuvant therapy were examined using multivariate logistic regression, controlling for clinicodemographic factors.

Results Among the 392 patients in the study cohort, 9.9% were diagnosed on screening examination, 53.1% underwent preoperative staging imaging and 74.5% underwent full preoperative colonic examination. On multivariate logistic regression, patients presenting via screening were more likely to move from presentation to diagnosis within the 4-week benchmark for this access time interval, compared with symptomatic patients (RR 8.1, P < 0.001). The absence of preoperative staging imaging was associated with achievement of the 4-week benchmark for the access time interval from diagnosis to surgery (RR 2.5, P < 0.001). Similarly, an absence of complete preoperative colonic examination was associated with achievement of the 8-week benchmark for the access time interval from surgery to adjuvant therapy (RR 6.6, P = 0.008).

Conclusions Although several associations between quality indicators and benchmarked access time intervals for colorectal cancer patients were identified, the relationship between quality and access is complex and far from universal. It is therefore clear that quality care and timely access are not synonymous, and that both must be studied to improve colorectal cancer care.

  • colorectal neoplasms
  • quality indicators
  • health care
  • health care quality
  • access and evaluation
  • cohort studies

Introduction

Colorectal cancer is a highly prevalent cancer, and remains a common cause of cancer death, among North American adults [1, 2]. The care of patients with colorectal cancer involves processes of screening, diagnosis, staging investigation, treatment and surveillance. It involves specialists in surgery, radiology, pathology, radiotherapy and chemotherapy, as well as primary care practitioners and many others. One of the greatest challenges of such a complex care pathway is the provision of high-quality care in a timely fashion.

Various measures of provider or health system performance, commonly referred to as quality indicators, are used to document key aspects of care, help set priorities for medical organizations, enhance accountability and contribute to quality-improvement efforts. In some jurisdictions, the performance of quality indicators has become a criterion for reimbursement of clinical services. The usefulness of a given quality indicator as a true measure of quality care is largely dependent on the level of evidence supporting its ability to predict patient outcomes. Ideally, quality indicators would be supported by randomized control trial data. However, the reality is that many quality indicators currently in use are based on expert opinion regarding standards of care and are not clearly linked to patient outcomes [3]. As with many other diseases, quality indicators specific to colorectal cancer care have been proposed by panels of experts for several key points in the decision and treatment algorithm of these patients [47].

Timely access to cancer care has become an important metric of health services delivery. In Canada, ‘wait-time guarantees’ exist as part of the national approach to cancer care. Other countries have established similar initiatives in response to public concern regarding timely access to care. Although there is a lack of conclusive evidence linking longer access times with worsened recurrence rates or survival, timely access is clearly linked to patient satisfaction [8, 9]. Published consensus statements provide broad descriptions of the components of timely access to cancer care and propose benchmarks for these access time intervals [1012]. To date there have been no published access time benchmarks specific to colorectal cancer, nor any specific data examining the applicability of generic cancer access time benchmarks to colorectal cancer.

At a time when emphasis on quality and timely access coexist in many jurisdictions, there is little data examining the relationship between these two elements of care. In particular, several of the proposed colorectal cancer-specific quality indicators necessitate investigations or consultations which have the potential to lengthen the time to diagnosis, definitive surgery and/or adjuvant therapy.

Thus, the objective of this study was to describe both the timeliness and quality of colorectal cancer care, and explore relationships between quality indicators and access time intervals for diagnosis, surgery, chemotherapy and radiation therapy.

Methods

From February 2002 to February 2004, all patients undergoing surgery for primary colorectal cancer within a single health district were enrolled in a prospective consecutive cohort study. This health district consists of an urban and semi-urban community of 400 000 people, serviced by one tertiary hospital system and one community hospital, with a total inpatient bed capacity of 1500. The tertiary hospital also provides referral-based services for a total catchment area of 132 000 km2 with a population of 1.8 million people.

Data on demographic and clinical factors, diagnostic and treatment events and pathology reporting were prospectively collected using a comprehensive, standardized method involving patient interviews, chart monitoring and survey administration [9]. IRB ethics approval was granted for this project (CDHA-RSA/2002-305).

Given the purpose of this study, patients who underwent surgery on an urgent basis after admission through the emergency department with a new diagnosis of colorectal cancer were excluded from the cohort used for analysis because the clinical status of these patients necessitated rapid action, outside the usual processes of referral, investigation and treatment. On the other hand, patients who initially presented to the emergency department and whose work-up lead to the diagnosis of colorectal cancer and surgery at a later date were included in the current analysis.

Quality indicators

The following colorectal cancer-specific quality indicators were selected a priori for examination in this study. They were selected because they are consistently proposed in the published literature [47], measurable through available healthcare databases and/or chart review, and were felt to be the quality indicators most likely to have an impact on timely access to care.

  1. Diagnosis made on screening examination (i.e., in an asymptomatic patient);

  2. Preoperative staging imaging of the liver (colon and rectal cancer) and pelvis (rectal cancer) by ultrasound, computed tomography (CT) or magnetic resonance imaging;

  3. Complete preoperative large-bowel examination by colonoscopy, or contrast enema plus flexible sigmoidoscopy;

  4. Pathology reporting that includes margin status;

  5. Among rectal cancer patients, a distal margin negative for malignant involvement;

  6. Pathology reporting that includes examination of at least 12 lymph nodes;

  7. In-hospital or 30-day mortality;

  8. Appropriate radiation oncology consultation, i.e. patient is seen:

    1. preoperatively for any rectal cancer patient;

    2. postoperatively for patients with Stage II or III rectal cancer;

  9. Appropriate medical oncology consultation, i.e. patient is seen:

    1. preoperatively or postoperatively for patients with Stage IV colon cancer;

    2. postoperatively for patients with Stage II or III colon cancer;

    3. preoperatively or postoperatively for patients with Stage II, III or IV rectal cancer.

The following definitions were used a priori to classify the available data regarding quality indicators. A diagnosis was considered to have been made on screening if the modality used to make the diagnosis was initiated without prior symptoms or signs suggestive of colorectal cancer. Preoperative colonic examination was considered complete if a full colonoscopy, or a combination of sigmoidoscopy and contrast enema, were performed. The presence of margin status data in the pathology report were dichotomously classified (present vs. absent). For the rectal cancer patients with a positive margin status in the original cohort database, original pathology reports were reviewed again to determine whether this consisted of a positive distal or radial margin.

The following quality indicators from the literature were not analysed in this study due to a lack of data: 5-year survival rates, local recurrence rates, anastomotic leak rates, total mesorectal type dissection rates, degree of nerve preservation and 1-year surveillance colonoscopy.

Access time intervals

Access time intervals for cancer care have been benchmarked in several jurisdictions [1012]. On the basis of the aggregate of existing benchmarks, we chose the following as benchmarks time intervals for this study: 4 weeks from presentation to definitive diagnosis; 4 weeks from diagnosis to surgery; and 8 weeks from surgery to adjuvant therapy.

In order to evaluate these time intervals within the studied cohort, several a priori definitions were necessary. Presentation date was defined as the date of the patient's first contact with the healthcare system directly addressing colorectal cancer presentation. For some, this was a screening physical exam including digital rectal exam, faecal occult blood testing or screening colonoscopy. For others, it was a physician visit regarding symptoms or an abnormal blood test result. The date of definitive diagnosis was the date on which definitive endoscopy or imaging studies were conducted. In cases where a suspicious barium enema was followed by confirmatory endoscopy, the date of the barium enema was used as the date of diagnosis. The start of chemotherapy or radiation was defined as the date of initial therapeutic administration. For rectal cancer patients who received preoperative radiotherapy, the interval from diagnosis to surgery was adjusted for ∼10 weeks of ‘therapy’ time to account for the delivery of radiation and an obligatory recovery time prior to surgery. This adjustment was made to provide a better estimate of the true system-related access time intervals for these patients.

Analysis

Several continuous variables were transformed into categorized variables. Age was categorized to identify the group of early cancers (<50 years of age—the age at which ‘average risk’ screening begins) and to divide the remaining cohort members into approximately equal-sized groups. The colorectal cancer tumour location was classified into three categories (right, left and rectum) based on the surgeon's description of location in dictated reports. Rectal cancer was defined as any lesion arising within 16 cm (6 in.) of the anal verge on endoscopy. Surgeons' colorectal cancer caseload was categorized around a threshold of 20 cases per year based on literature that indicates this is a reasonable yearly caseload to maintain surgical skill in complex colorectal cases [13, 14]. Charlson Comorbidity Index was categorized using a standard cut-off of 5, defining groups of patients with ‘mild’ versus ‘moderate-severe’ comorbidities [15]. Postoperative length of stay (LOS) was categorized using a cut-off value of the 75th percentile for days spent in hospital.

The proportion of patients seen appropriately by medical or radiation oncologists was calculated with a denominator including all patients who qualified for consultation according to the criteria listed above, and a numerator consisting only of those patients who were actually seen.

Descriptive statistics were generated regarding clinicodemographic factors, performance of quality indicators and achievement of benchmarked access time intervals between presentation, diagnosis, surgery and adjuvant therapy. Relationships between factors in each of these three data groups were first analysed using univariate logistic regression. The results of this univariate analysis are not presented in this paper but all factors were included in the subsequent multivariate analyses.

The first part of the multivariate logistic regression was to analyse associations between clinicodemographic factors and performance of quality indicators. For quality indicators pertaining to diagnosis, staging, pathology and mortality, the following factors were analysed: age, gender, body mass index (BMI), Charlson Comorbidity Index, surgeon caseload, hospital type, presentation type, tumour location and TNM stage. For the quality indicators of medical and radiation oncology consultation, LOS and preoperative radiotherapy were added to the model. Due to redundancy of data or obvious interactions, particular variables were excluded from particular analyses:

  1. Presentation type was excluded from all analyses of the ‘Diagnosis made on screening’ quality indicator.

  2. Given that the definition of appropriate medical oncology consultation was inherently dependent on TNM stage, this factor was not included in analysis of this quality indicator.

  3. Given that the definition of appropriate radiation oncology consultation was inherently dependent on tumour location and TNM stage, these factors were not included in analysis of this quality indicator.

Finally, each benchmarked access time interval was analysed using multivariate logistic regression with clinicodemographic factors and quality indicators included as categorical variables. For the access time interval presentation-to-diagnosis, factors included in the analysis were: age, gender, BMI, Charlson Comorbidity Index, surgeon caseload, hospital type, tumour location, TNM stage and the quality indicators of diagnosis made on screening, preoperative staging imaging, preoperative full colon examination and perioperative mortality. For the access time interval diagnosis-to-surgery, all of these factors were included in the analysis, plus preoperative radiotherapy.

For the access time interval surgery-to-adjuvant therapy, multivariate analysis incorporated all of these factors plus LOS, and the quality indicators pertaining to pathology reporting. The quality indicators regarding negative distal margin status was excluded due to the very small number of patients [2] affected by this criteria. Quality indicators regarding appropriate medical and radiation oncology consultation were not included as all patients who received adjuvant therapy had been seen by the corresponding oncologist.

For access time intervals presentation-to-diagnosis and diagnosis-to-surgery, additional multivariate analyses were performed, including only patients who qualified for medical or radiation oncology consultation by the criteria outlined earlier. In these analyses, the quality indicators pertaining to appropriate medical or radiation oncology were included in the multivariate model.

All analyses were performed using SPSS 15.0 (Chicago, IL, USA).

Results

Demographics

There were 455 individuals included in the prospective database of patients undergoing colorectal cancer surgery within the study time period. Sixty-three of these underwent urgent surgery after having presented to the emergency department with severe symptoms, and were excluded, resulting in a final cohort of 392 patients.

Demographic characteristics of the study cohort are summarized in Table 1. Of note, 49% of patients were over 70 years of age, 34% were obese (BMI ≥ 30 kg/m2) and 19% had moderate-severe comorbidities using the Charlson Comorbidity Index.

View this table:
Table 1

Demographics (n = 392)

n%
Age<50 years297.4
50–69 years17243.9
≥70 years19148.7
GenderMale21254.1
Female18045.9
BMIa (kg/m2)Normal/underweight (0–24.9)10627.0
Overweight (25.0–29.9)14637.2
Obese (≥30.0)13434.2
Charlson Comorbidity IndexMild (1–4)31881.1
Moderate–Severe (≥5)7418.9
Hospital typeTertiary care centre29274.5
Community hospital10025.5
Surgeon caseload<20 cases/year20652.6
>20 cases/year18647.4
Colorectal cancer
 Presentation typeScreening/asymptomatic399.9
Symptomatic35390.1
 Tumour locationRight colon15138.5
Left colon10827.6
Rectum13333.9
 TNM StageI9724.7
II11930.4
III13033.2
IV4611.7
 LOS≤75th percentile (12 days)29675.5
>75th percentile9624.5
  • BMI, Body Mass Index; LOS, Length of Stay.

  • aMissing data points result in category total less than the total sample size.

Quality indicators

Table 2 summarizes the performance of quality indicators within the study cohort. Most notably, preoperative staging imaging was completed in only 53% of patients, preoperative colonic examination in 75% and ≥12 lymph nodes were examined in only 45% of patients.

View this table:
Table 2

Performance of quality indicators (n = 392)

n%
Diagnosis made on screening399.9
Preoperative imaging (liver, pelvis)20853.1
Preoperative full colon examination29374.7
Pathology report that includes margin status38698.5
Negative distal margina131/13398.5
Pathology report examines at least 12 lymph nodes17845.4
In-hospital/30-day mortality153.8
Appropriate radiation oncology consultationb75/11963.0
Appropriate medical oncology consultationc163/29555.3
  • aAmongst rectal cancer patients (n = 133).

  • bAmongst rectal cancer patients with TNM stage II or III disease (n = 119).

  • cAmongst rectal and colon cancer patients with TNM stage II, III or IV disease (n = 295).

Results of the multivariate analysis of factors associated with the performance of quality indicators are contained in Table 3. Colorectal cancers diagnosed on screening examination tended to be earlier stage tumours and to occur in younger patients. Preoperative staging imaging was completed more often for patients with rectal cancer and those with advanced disease, and was also completed in the community hospital more often than in the tertiary care hospital. Preoperative full colon examination was completed most often in patients cared for by surgeons with a high colorectal cancer case load and least often in patients with rectal cancer. In terms of the pathological examination of ≥12 lymph nodes, there was significant disparity between the tertiary and community hospitals. Appropriate radiation and medical oncology consultation appeared to occur less frequently among older patients, females and those with morbid obesity.

View this table:
Table 3

Multivariate analysesa: quality indicators (n = 392)

Quality indicatorbSignificant factorsnPerformance of quality indicator (%)RRP-value
Diagnosis made on screeningAgeOverall0.1
≥701918.61.0
50–6917210.51.40.4
<502917.93.80.04
Surgeon caseload<20/year2067.51.0
≥20/year18613.03.00.04
TNM stageOverall0.03
IV462.21.0
III13010.28.30.06
II1197.75.40.1
I9716.515.70.01
Preoperative imaging (liver, pelvis)Hospital typeTertiary care Community29249.71.0
10063.01.90.04
Tumour locationOverall<0.001
Right15140.11.0
Left10852.11.80.04
Rectum13370.54.4<0.001
TNM stageOverall0.2
I9749.01.0
II11955.82.00.03
III13050.81.40.2
IV4660.91.80.2
Preoperative full colon examinationSurgeon caseload<20/year20667.01.0
≥20/year18684.32.30.008
Charlson1–431772.91.0
Comorbidity Index≥57485.52.40.02
Tumour locationOverall0.002
Rectum13366.71.0
Left10886.53.20.001
Right15175.52.20.009
Pathology report examines at least 12 lymph nodesGenderMale21238.71.0
Female18053.31.70.04
Hospital typeCommunity1008.51.0
Tertiary29258.322.0<0.001
Tumour locationOverall0.03
Rectum13341.11.0
Left10838.51.00.9
Right15154.42.10.02
TNM stageOverall0.02
I9733.01.0
II11947.82.20.02
III13052.52.50.004
IV4648.82.80.03
Appropriate radiation oncology consultation (n = 119)GenderMale6771.61.0
Female5251.90.30.03
BMIOverall0.04
≥30.04151.21.0
25.0–29.94770.24.30.03
0–24.93167.73.50.03
Appropriate medical oncology consultation (n = 296)AgeOverall<0.001
≥7014934.91.0
50–6912373.23.9<0.001
<502487.57.20.004
Tumour locationOverall0.006
Right12840.61.0
Left8055.01.60.2
Rectum8876.13.10.001
  • aLogistic regression with demographic and clinical factors as categorical variables. Presentation type was excluded from all analyses of the ‘Diagnosis made on screening’ quality indicator. Tumour location and TNM stage were excluded from analysis for the radiation oncology quality indicator. TNM stage excluded for analysis of the medical oncology quality indicator. bThe following QIs had no significant associations with clinicodemographic factors on multivariate analysis and are not shown: ‘Pathology report contains margin status’, ‘In-hospital or 30-day mortality’, ‘Negative distal margin (rectal cancer patients only)’.

Multivariate analysis of the quality indicators regarding pathology reporting of margin status, negative distal margin status among rectal cancer patients and perioperative mortality did not reveal any associated clinicodemographic factors.

Access time intervals

Median time intervals for access to diagnosis, surgery, and adjuvant therapy are summarized in Table 4. The achievement of benchmarked access time intervals ranged from 35 to 51% with a median time interval from presentation to diagnosis of 51 days, a median time interval between definitive diagnosis and surgery of 28 days, and a median time interval from surgery to adjuvant therapy of 60 days.

View this table:
Table 4

Achievement of benchmarks for access time intervals (n = 392)

nMedian time (days)Benchmark access time (weeks)Achievement of benchmark access time (%)
Presentation to diagnosis39251434.9
Diagnosis to surgery39228a451.3
Surgery to adjuvant therapy12260845.9
  • aIncorporates 10-week adjustment for rectal cancer patients who received preoperative radiation (see text for details).

On multivariate analysis (Table 5), the access time interval from presentation to diagnosis was shorter among those who had a diagnosis made on screening and those who suffered perioperative mortality. Achievement of the benchmark for the diagnosis-to-surgery access time interval was associated with a lack of preoperative staging imaging, colonic rather than rectal lesions, more advanced disease stage and, among rectal cancer patients, with not having had preoperative radiotherapy. The benchmark for time from surgery to adjuvant therapy, among appropriate patients, was achieved more often in those patients who had not undergone preoperative full colonic examination and those who did not have a prolonged LOS in hospital postoperatively.

View this table:
Table 5

Multivariate analysesa: access time intervals (n = 392)

Access time intervalSignificant factorsnAchievement of benchmark access time (%)RRP-value
Presentation to diagnosis: benchmark 4 weeksDiagnosis made on screeningNo35330.01.0
Yes3974.48.1<0.001
In-hospital or 30-day mortalityNo37734.51.0
Yes1561.53.80.03
Diagnosis to surgery: benchmark 4 weeksbPreoperative imaging (liver, pelvis)Yes20839.51.0
No18464.02.5<0.001
Tumour locationOverall0.002
Rectum13329.31.0
Left10860.22.60.003
Right15164.22.60.002
TNM stageOverall0.001
I9733.01.0
II11952.92.20.02
III13056.92.80.001
IV4669.65.10.001
Preoperative radiation therapyYes306.71.0
No36255.05.20.04
Surgery to adjuvant therapy: benchmark 8 weeks (n = 122)Preoperative full colon examinationYes9138.51.0
No3167.76.60.008
Length of stay>75th percentile1216.71.0
≤75th percentile11049.17.80.02
  • aLogistic regression with demographic factors and quality indicators included as categorical variables. The endpoint access time intervals were categorized by published benchmarks. Presentation type was excluded from all analyses because the QI. ‘Diagnosis made on screening’ was included in all. bRectal cancer patients who received preoperative radiation therapy had their ‘Definitive diagnosis to surgery’ access time adjusted for a presumed 4 weeks of radiation therapy and 6 weeks of mandatory waiting time between completion of radiation therapy and surgery. If their total wait time was less than 10 weeks (i.e. they never completed radiation therapy) their original total access time was used.

Discussion

Demographics

The cohort of patients studied here are typical of current trends in North American healthcare—they tended to be older, obese and have multiple co-morbidities. These factors were associated with the performance of several quality indicators but did not appear to be related to achievement of benchmarked access time intervals. Other clinicodemographic factors, such as distance from home to hospital, education level and frequency of visits to a primary care physician, may be more closely related to access time intervals but were not captured by the prospective data collection methodology used in the current study.

Quality indicators

High-quality care is the primary goal of health-care professionals, administrators and policy-makers. The Institute of Medicine has defined ‘quality’ as it pertains to healthcare as the extent to which health services for individuals and populations increase the possibility of desired health outcomes and are consistent with current professional knowledge [7]. Measuring quality requires indicators that are reliably measureable, consistently defined and (ideally) closely linked to improved patient outcomes. In general, the available quality indicators for colorectal cancer care, as well as other types of cancer, have been defined by various review committees and, although prospective healthcare databases are being developed and used more widely, are not yet based on prospective data. We were able to gather quality indicator information on 100% of the current cohort but performance of quality indicators among the cohort was highly variable, ranging from 10 to 98%. This is similar to the findings of the National Initiative for Cancer Care Quality in the USA, who found 50–93% performance of quality indicators among 478 colorectal cancer patients [16].

The low rate of diagnosis on screening examination (9.9%) is similar to rates in the literature [17]. Organized colorectal cancer screening programmes have been shown to increase the number of colorectal cancer diagnoses and decrease the stage of colorectal cancer on diagnosis [18] but no such programme was in effect at the time of this study.

The rate of preoperative staging imaging in this cohort (53%) is below that reported by others [19]. In this study cohort, those cared for in the community hospital were more likely to have preoperative staging imaging than those cared for in the tertiary care hospital—perhaps indicating a local access issue. Not surprisingly, patients with rectal tumours and higher stage lesions were also more likely to have undergone preoperative staging imaging.

Preoperative colonic examination was completed in 75% of patients, with no comparable values available in the literature. On multivariate analysis, patients who were cared for by a high-volume surgeon were more likely to have undergone preoperative colonic examination—perhaps reflecting easier access to endoscopic services. Patients with a higher comorbidity index were more likely, whereas those with rectal lesions were less likely, to have undergone colonic examination. There is no clear explanation for these findings, although it may be hypothesized that a surgeon would be more interested in ruling out synchronous lesions preoperatively in a patient with severe comorbidities and that endoscopists may hesitate to perform complete bowel preparation and examination in larger rectal tumours.

Examination of at least 12 lymph nodes was included in the pathology report of only 45% of this cohort's patients. This is comparable to the 37% adequate lymph node evaluation rate reported by Baxter et al. [20]. The perioperative mortality rate in this cohort (3.8%) is also comparable to that reported for open colectomy in recent clinical trials, which ranges from 0 to 4.9% [21].

Of patients who qualified for medical oncology consultation in the current cohort, only 55% (163/296) were seen by a medical oncologist and 37.5% (111/296) received chemotherapy. This compares somewhat unfavourably with the literature where 70% of eligible patients were seen by a medical oncologist and 40–66% of eligible patients received adjuvant chemotherapy [2226]. In terms of radiation oncology consultation, 63% (75/119) of patients in this cohort who qualified were seen by a radiation oncologist. This compares with the literature where 44–64% of eligible patients received radiation therapy [22, 26, 27]. On multivariate analysis of our cohort, medical oncology consultation was significantly related to younger age and rectal tumour location. Appropriate radiation oncology consultation was less likely among females and patients with morbid obesity. Further study is required to elucidate the barriers to medical and/or radiation oncology consultation and receipt of adjuvant therapy in colorectal cancer patients.

Access time intervals

Overall, only 35% of patients achieved the 4-week benchmark for the access time interval from presentation to diagnosis. Over 30% of patients in this study waited more than 3 months between documented presentation (with symptoms or abnormal findings on physical examination or laboratory investigation) and definitive diagnosis on imaging or endoscopy. This ‘pre-diagnosis lag time’ is recognized as one of the most difficult to address because it involves many aspects of the healthcare system—primary care physicians, laboratories, endoscopy resources, radiology resources, etc. Given the non-specific nature of colorectal cancer symptoms, many patients must be investigated for each cancer found. For instance, only 6% of patients presenting with iron-deficiency anemia had a colorectal cancer on investigation [28]. In response, several jurisdictions have developed referral guidelines and target wait-times, as well as continuous monitoring and reporting of benchmark achievement such as is being done by Cancer Care Ontario (website: www.csqi.cancercare.on.ca). However, several authors have found that a significant proportion of patients diagnosed with cancer did not fall within these guidelines, nor were they provided care within the target access time [29, 30].

The finding that patients with rectal cancer were less likely to achieve the 4-week benchmark for diagnosis-to-surgery than those with colon cancer is concerning, particularly because it was independent of the factors of preoperative staging workup and preoperative radiation therapy. Similar findings—that there are longer delays in the treatment of rectal cancer patients—have recently been reported from Denmark [31] but the patient, physician and systemic factors contributing to this finding have not yet been clearly delineated.

Preoperative staging imaging was associated with a reduced likelihood of meeting the 4-week benchmark from diagnosis-to-surgery. To our knowledge this finding has never been previously reported. In discussion with surgeons in our district and others across Canada, this finding reflects their experience, particularly in the earlier days of CT scanning—that the usefulness of preoperative imaging must be balanced against the delay it can create in surgical booking. There may be other factors contributing to a delay in the diagnosis-to-surgery time interval that have not been captured by the current data set. These may include variation in length of waiting lists, inpatient bed availability and mechanisms for preoperative assessment by anesthesia specialists, among others. Further study and streamlining of such factors is urgently needed to improve access times.

Preoperative radiation therapy was associated with a longer access time interval from diagnosis to surgery, even after adjusting for 10 weeks of therapeutic time. On closer examination of the data for the 30 patients who received preoperative radiation, 21 (70%) waited more than 4 weeks from diagnosis to start of preoperative therapy. Only 4 (13%) waited more than 4 weeks from the end of the 10-week preoperative therapy time till undergoing surgery. This suggests that the majority of the delay between diagnosis and surgery in these patients occurs whereas they wait for preoperative therapy. This finding is similar to that of a recent study from the UK, where a delay in starting preoperative chemoradiation was the most common reason for failure to achieve target wait-times for surgery [30].

The independent relationship between preoperative full colonic examination and the access time interval from surgery to adjuvant therapy is difficult to explain. There must be factors, perhaps unique to the healthcare region, which are influencing this relationship and have not yet been identified.

The limitations of the current study include its modest sample size, use of data from only one geographic region, and inability to examine several important quality indicators due to a lack of data. To this end, we are currently examining the relationships between quality indicators and access time intervals in the entire province of Nova Scotia using administrative and cancer registry data. However, collection of specific quality indicator and access time interval data in a larger multicentre prospective study could better answer questions such as:

  1. Is the presence of a structured colorectal cancer screening programme associated with improved performance of quality indicators and achievement of benchmarked access time intervals?

  2. Is the availability of imaging facilities truly a significant factor in achievement of benchmarked access time intervals from diagnosis to surgery?

  3. What factors act as barriers to the performance of preoperative staging imaging and colonic examination?

  4. Can the finding of a longer delay to treatment among rectal cancer patients be generalized to other healthcare systems and, if so, what factors contribute to this delay?

  5. Most importantly, are there associations between quality indicators, access time intervals and outcomes such as cancer recurrence and 5-year survival?

For the cohort included in this study, follow-up data including recurrence and survival outcomes are currently being collected and will be the subject of future publications.

Finally, patients' perceptions regarding the quality of cancer care and access to cancer treatment may influence patient satisfaction more than objectively measured quality indicators and access time intervals [32]. Data regarding such perceptions were not collected in the current study but should be further investigated to better define quality, timely cancer care.

Conclusion

In summary, there was variable performance of quality indicators specific to colorectal cancer among this cohort of patients. The majority of patients failed to achieve benchmarked time intervals for access to care. The relationship between achievement of quality indicators and benchmarked time intervals was complex and multifactorial. Patients achieving the 4-week benchmark from presentation to diagnosis were most likely to have had their diagnosis made on screening examination. Patients achieving the 4-week benchmark from diagnosis to surgery were less likely to have undergone preoperative staging imaging of their liver and/or pelvis. Patients achieving the 8-week benchmark from surgery to start of adjuvant chemotherapy were less likely to have undergone preoperative full colonic examination. It is therefore clear that quality care and timely access are not synonymous, and that both must be studied to improve colorectal cancer care.

Funding

Nova Scotia Health Research Foundation.

References

View Abstract