SEARCH

SEARCH BY CITATION

Keywords:

  • cancer survivors;
  • breast cancer;
  • colorectal cancer;
  • guideline concordance

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

BACKGROUND:

Evidence-based guidelines recommend routine surveillance, including office visits and testing, to detect new and recurrent disease among survivors of breast and colorectal cancer. The extent to which surveillance practice is consistent with guideline recommendations or may vary by age is not known.

METHODS:

Cohorts of adult patients diagnosed with breast (n = 6205) and colorectal (n = 2297) cancer between 2000 and 2008 and treated with curative intent in 4 geographically diverse managed care environments were identified via tumor registries. Kaplan-Meier estimates were used to describe time to initial and subsequent receipt of surveillance services. Cox proportional hazards models evaluated the relation between patient characteristics and receipt of metastatic screening.

RESULTS:

Within 18 months of treatment, 87.2% of breast cancer survivors received recommended mammograms, with significantly higher rates noted for patients aged 50 years to 65 years. Among survivors of colorectal cancer, only 55.0% received recommended colon examinations, with significantly lower rates for those aged ≥ 75 years. The majority of breast (64.7%) and colorectal (73.3%) cancer survivors received nonrecommended metastatic disease testing. In patients with breast cancer, factors associated with metastatic disease testing include white race (hazards ratio [HR], 1.13), comorbidities (HR, 1.17), and younger age (HR, 1.13; 1.15; 1.13 for age groups: <50, 50–64, and 65–74 respectively). In those with colorectal cancer, these factors included younger age (HR, 1.31; 1.25 for age groups: <50 and 50–64 respectively) and comorbidities (HR, 1.10).

CONCLUSIONS:

Among an insured population, wide variation regarding the use of surveillance care was found by age and relative to guideline recommendations. Breast cancer survivors were found to have high rates of both guideline-recommended recurrence testing and non–guideline-recommended metastatic testing. Only approximately 50% of colorectal cancer survivors received recommended tests but greater than 67% received metastatic testing. Cancer 2012. © 2012 American Cancer Society.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

In 2006, the Institute of Medicine (IOM) recommended that cancer survivors receive ongoing surveillance care based on a follow-up plan that is clearly and effectively explained.1 Evidence-based and consensus-based guidelines from the National Comprehensive Cancer Network (NCCN),2, 3 the American Society of Clinical Oncology (ASCO),4, 5 and other specialty-oriented medical societies6-9 outline recommended schedules for ongoing surveillance care after cancer treatment with curative intent. Although recommendations vary to some extent in terms of the type and timing of care, NCCN and ASCO guidelines for survivors of breast and colorectal cancer recommend a course that is comprised of periodic physical examinations and screening for recurrent and metastatic disease via imaging and other procedures.

To the best of our knowledge, relatively little is known regarding the quality of care delivered to cancer survivors. Prior studies among survivors of breast and colorectal cancer have found deviations in surveillance care patterns relative to evidence-based guidelines as well as variations in surveillance care use by patient sociodemographic characteristics, including race and income. However, such studies have been limited to survivors receiving care from 1 delivery organization10-16 or those enrolled in Medicare.17-21 As such, the care received by survivors who are aged ≤ 64 years is not well documented. Furthermore, to the best of our knowledge, no studies have been conducted since the IOM report was issued.

Using cohorts of breast and colorectal cancer survivors receiving care from 4 geographically diverse health maintenance organizations between 2000 and 2008, we evaluated the extent to which surveillance care use was consistent with guideline recommendations. Of specific interest was the evaluation of the variability in surveillance care use by age at diagnosis.

MATERIALS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Study Population and Setting

This study was conducted within 4 large, nonprofit, integrated health systems: Group Health Cooperative (Seattle, Wash), Health Alliance Plan/Henry Ford Health System (Detroit, Mich), Kaiser Permanente Colorado (Denver, Colo), and Kaiser Permanente Northwest (Portland, Ore). These plans are all members of the Cancer Research Network (National Cancer Institute [NCI] Cooperative Agreement No. U19 CA79689, Increasing Effectiveness of Cancer Control Interventions), an initiative of the NCI designed to conduct research on cancer prevention, early detection, treatment, long-term care, surveillance, and cancer communication and dissemination and implementation research.

Data available from each organization's tumor registry were used to identify patients aged ≥ 18 years who were diagnosed between January 1, 2000 and December 31, 2008 with nonmetastatic breast (International Classification for Diseases for Oncology [ICD-O] codes C50.0-C50.9) or colorectal (ICD-O codes C18.0-C18.9, C19.9, C20.9, and C26.0) cancer. Patients eligible for study inclusion were those continuously enrolled in the health plan for the 1-year period preceding their date of cancer diagnosis. Patients for whom no information regarding stage of disease was available at the time of diagnosis were excluded. Patients were also excluded if they had a previous diagnosis of invasive cancer, had distant metastases at the time of diagnosis, or did not receive treatment with curative intent (ie, surgery). The breast cancer cohort was limited to females and excluded women who had undergone a bilateral mastectomy or 2 separate occurrences of a unilateral mastectomy on different dates.

Sample patients were followed from an “index date,” defined as 3 months after the last curative surgical procedure, to the first of the following endpoints (“end date”): death, tumor recurrence, diagnosis of a second primary tumor, health plan disenrollment 5 years after the initial cancer diagnosis, or the end of the follow-up period (December 31, 2008). Previous findings indicate that altering assignment of the index date to exclude from consideration care received within the first 6 and 9 months after surgical treatment (vs 3 months) does not alter results or conclusions.13 Institutional Review Boards at each of the 4 participating organizations approved all aspects of the study protocol.

Data Sources and Measures

Automated electronic medical record (EMR) and claims data linked to tumor registry data were accessed to obtain patient demographic characteristics, date of cancer diagnosis, stage at diagnosis, and comorbidities within the 12-month period preceding diagnosis for each patient. Patient demographics included age at diagnosis, gender, and race (black, white, and other). Using the patient's residential street address combined with 2000 US Census tract data, we estimated the median household income for each patient's residential neighborhood. Clinical measures included stage of disease at diagnosis, date of any cancer recurrence(s) (defined as either local recurrence or the occurrence of a second primary tumor) within 5 years of the index date, and date of death. We measured recurrence(s) using the algorithm for measuring disease-free survival described in Lamont et al.22 We summarized the disease stages into 3 groups: in situ (American Joint Committee on Cancer [AJCC] general stage 0), localized (AJCC general stage 1), and regional (AJCC general stages 2-5 [ie, regional by direct extension, lymph nodes, both, or not otherwise specified]).

In addition to data available via the tumor registries, EMR and claims data were reviewed between the patient's “index date” and “end date.” For each organization, these data sources contain comprehensive visit and procedure information for any outpatient encounter to a physician, laboratory, and imaging department. These data were used to compute receipt of testing services and treatment with surgery, chemotherapy, hormone therapy, radiation, and combination therapies. The Deyo et al adaptation of the Charlson comorbidity index and each of its component diagnostic subgroups was constructed using inpatient and outpatient diagnostic information available in the 12-month period preceding diagnosis.23

Surveillance Care Receipt

The primary outcome of interest is the time to receipt of recommended surveillance care. For both breast and colorectal cancer survivors, we observed 3 distinct types of surveillance: physical examinations, testing for local recurrence, and testing for metastatic disease. Physical examinations performed by a primary care provider, medical oncologist, radiation oncologist, general surgeon, or gastroenterologist (for patients with colorectal cancer only) were included. For those with breast cancer, local recurrence tests included mammography, magnetic resonance imaging (MRI), and ultrasound, whereas for patients with colorectal cancer, local recurrence tests included colonoscopy, sigmoidoscopy, and barium enema. Procedures to detect metastatic disease recurrence included chest radiograph; computed tomography of the chest, abdomen, pelvis, or head; MRI of the chest, abdomen, pelvis, or head; bone scan; gallium scan; liver/spleen scan; and abdominal or pelvic ultrasound. Because surveillance care guidelines recommend not only the type of care but also its timing, we evaluated the frequency of care receipt and the time interval between the initial and subsequent receipt of the same type of examination/test.

Finally, we evaluated whether the patient received NCCN and ASCO guideline-recommended surveillance care. This was done by estimating the time (in days) from the index date to the date by which a patient had received the minimum amount and type of surveillance testing recommended by the guideline. For patients with breast cancer, this included 2 physical examinations and 1 mammogram, whereas for those with colorectal cancer, this included 2 physical examinations and 1 complete examination of the colon. In addition to time, we also reported the percentage of patients who received recommended care within 18 months after treatment with curative intent. This was done because care receipt in practice rarely falls neatly within 12-month periods.

Analytical Approach

Because of the differing lengths of follow-up among sample members, we used Kaplan-Meier estimates to evaluate the median time to initial and subsequent care receipt by type of examination/test. We reported the percentage of patients receiving care within the corresponding 18 months using actuarial results. Kaplan-Meier estimates were also used to determine cumulative incidences of the receipt of the minimum recommended surveillance care. Estimates were evaluated for the entire population of survivors as well as separately for 4 age groups: those aged < 50 years, those aged 50 years to 64 years, those aged 65 years to 74 years, and those aged ≥ 75 years. We plotted values as the probability of the complement so that graphs reflected time to receipt of care. We used Cox proportional hazards models, which also account for the differing length of follow-up among sample members, to quantify the effects of baseline clinical and sociodemographic patient characteristics, including patient age group, on the risk of receiving metastatic testing (a nonrecommended care type) within the initial 18 months. In addition to age, models were adjusted for the patient's race, sex (only for the colorectal cancer sample), stage, neighborhood household income, Charlson-Deyo comorbidity index, and health plan. All analyses were performed using SAS software (version 9.2; SAS Institute Inc, Cary, NC). Statistical results achieving a P < .05 were considered to be statistically significant.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Cohort Characteristics

The characteristics of the 6205 breast cancer patients and 2297 colorectal cancer patients by health plan are shown in Table 1. The mean ages of the cohorts were 62.4 years (± 12.6 years) for patients with breast cancer and 68.6 years (± 12.2 years) for patients with colorectal cancer, with 57% of the breast cancer patients aged < 65 years and 35% of the patients with colorectal cancer aged < 65 years. Although the study included only female breast cancer patients, the gender distribution for colorectal cancer was 49% male and 51% female. African Americans represented 8% of the breast cancer sample and 8% of the colorectal cancer sample. For breast cancer, the majority of tumors were classified as being of local stage (80%), whereas among the colorectal cancer group there was a more even distribution between patients with local (49%) and regional disease (51%). Per study eligibility criteria, all patients received curative surgical treatment. The overall median length of observation was 49.5 months for patients with breast cancer and 40.9 months for patients with colorectal cancer. Evidence of recurrent cancer or a second primary tumor occurred in 13% of patients with breast cancer and 22% of patients with colorectal cancer. Nearly 96% of patients with breast cancer survived through the end of follow-up, compared with 90% of patients with colorectal cancer. This survival rate was 99% for those with breast cancer who were aged < 65 years versus 91% for patients aged ≥ 65 years. In patients with colorectal cancer, the survival rate for patients aged < 65 years was 97% compared with 86% for patients aged ≥ 65 years. Survival differences by age group (aged < 65 years vs aged > 65 years) in both the breast and colorectal cancer patient cohorts were statistically significant (data not shown).

Table 1. Sample Characteristics by Cancer Site
CharacteristicBreast n = 6205Colorectal n = 2297
  1. Abbreviation: SD, standard deviation.

Mean age at diagnosis (SD), y62.4 (12.6)68.6 (12.2)
Age at diagnosis, %  
 <50 y166
 50-64 y4129
 65-74 y2429
 ≥75 y1936
Sex, %  
 Male49
 Female10051
Race, %  
 Black88
 White8181
 Other1111
Median household income (SD)$53,906 ($20,354)$51,580 ($19,237)
Stage, %  
 In situ244
 Localized5645
 Regional2051
Site of care, %  
 Health plan 12218
 Health plan 23132
 Health plan 31918
 Health plan 42832
Treatment, %  
 Surgery100100
 Chemotherapy2933
 Hormone therapy49
 Radiotherapy6112
 Combination treatments7033
Median follow-up, mo49.540.9
Recurred within 5 y, %1322
Survived 5 y, %9690

Receipt of Surveillance Care

Results from the Kaplan-Meier survival analysis for the time to initial and subsequent receipt of specific examinations and procedures are shown in Table 2. These results are presented for all ages combined, and separately for each of the 4 aforementioned age groups. For both cancer sites, the overwhelming majority of patients (97.1% of patients with breast cancer and 98.2% of patients with colorectal cancer) received the minimum recommended number of physical examinations within the initial 18 months after treatment with curative intent. The median time to receipt of the recommended physical examinations ranged from 1.6 months among patients with breast cancer to 1.8 months among patients with colorectal cancer. If we consider the time to subsequent receipt of physical examinations, then a majority of cancer survivors received at least twice the number of recommended physical examinations within the initial year after treatment. Among both breast and colorectal cancer patients, the median time to receipt of a subsequent physical examination was just over 2 months, suggesting that many patients received 2 physical examinations within the initial 4 months after treatment.

Table 2. Time to Receipt of Physical Examinations and Local Recurrence and Metastatic Disease Testing by Service Type and Age Among Breast (n=6205) and Colorectal (n=2297) Cancer Survivors
SurvivorAge Group, Years
Median Time to Initial Receipt, MonthsAll<5050-6465-74≥75
  • a

    When sample size did not permit estimation of the median time, the number of cohort members receiving the service of interest during the follow-up period is reported.

Breast cancerTwo physical examinations1.61.21.51.72.1
 One mammogram7.27.97.06.67.8
 Any metastatic disease testing9.49.19.58.89.9
Colorectal cancerTwo physical examinations1.81.61.71.81.9
 One complete examination of colon13.310.510.711.528.8
 Any metastatic disease testing6.14.36.15.96.9
Percentage of Patients in Receipt Within 18 Months of TreatmentAll<5050-6465-74≥75
Breast cancerTwo physical examinations97.196.396.698.197.6
 One mammogram87.281.689.391.182.5
 Any metastatic disease testing64.761.663.467.467.1
Colorectal cancerTwo physical examinations98.298.697.098.699.2
 One complete examination of colon55.063.064.162.141.0
 Any metastatic disease testing73.373.372.973.274.8
Median Time Between Initial and Subsequent Receipt, MonthsAll<5050-6465-74≥75
Breast cancerTwo physical examinations2.11.62.12.12.4
 One mammogram12.312.412.212.212.6
 Any metastatic disease testing15.916.821.013.813.8
Colorectal cancerTwo physical examinations2.32.12.12.12.6
 One complete examination of colon(n=412)a(n=34)a(n=103)a(n=111)a(n=86)a
 Any metastatic disease testing7.46.98.47.57.1
Percentage of Patients in Receipt Within 18 Months of Initial ReceiptAll<5050-6465-74≥75
Breast cancerTwo physical examinations94.393.293.296.695.1
 One mammogram75.469.479.080.566.7
 Any metastatic disease testing51.350.848.653.555.3
Colorectal cancerTwo physical examinations90.892.086.793.093.0
 One complete examination of colon16.726.417.819.812.2
 Any metastatic disease testing62.562.160.962.464.7

Most breast cancer patients received recommended local recurrence testing within 18 months. Among this group, 87.2% of patients underwent mammograms within 18 months of treatment, with a median of 7.2 months. Only 55% of colorectal cancer survivors received a complete examination of the colon within the first 18 months. The median time to initial receipt of this test was 13.3 months. Table 2 also shows the relatively common use of metastatic disease testing among cancer survivors. Nearly 65% of breast cancer survivors and 73% of colorectal cancer survivors received some type of metastatic disease testing within 18 months after treatment with curative intent. The median time to the initial receipt of metastatic testing in patients with breast cancer was 9.4 months, and was 6.1 months in patients with colorectal cancer.

When comparing the 4 age groups among breast cancer survivors, the percentage of patients receiving a mammogram within 18 months of treatment with curative intent was similar among the middle 2 age groups, with 89% and 91%, respectively, but was significantly lower in the group of patients aged < 50 years (82%) and those aged ≥ 75 years (82%). Among colorectal cancer survivors, the younger 3 age groups had similar rates of colon examinations within 18 months of treatment with curative intent, with 63%, 64%, and 62%, respectively, but a significantly lower rate was noted among the survivors aged ≥ 75 years (41%). In the interim,, we observed a small but significant increase in the receipt of metastatic testing within the first 18 months after treatment with curative intent associated with increasing age for both breast and colorectal cancer survivors.

Figure 1 shows estimates of the time to minimum guideline-recommended surveillance care during the initial year (ie, 2 physical examinations and 1 mammogram for breast cancer survivors and 2 physical examinations and 1 complete examination of the colon for colorectal cancer survivors), stratified by age group. Overall, nearly 90% of breast cancer survivors received the recommended surveillance care within 18 months of treatment with curative intent, although the rates are higher for those patients aged 50 years through 74 years compared with their younger and older counterparts. It is interesting to note that, regardless of age, the overwhelming majority of these patients received that care well before reaching the 12-month mark after their treatment with curative intent. Conversely, just greater than one-half of colorectal cancer survivors received the recommended surveillance care within 18 months of treatment with curative intent, with all age groups having similar rates except for the patients aged ≥ 75 years, who recorded significantly lower rates of receipt.

thumbnail image

Figure 1. The likelihood of initial 1-year minimum guideline-recommended service receipt for survivors of breast and colorectal cancer is shown.

Download figure to PowerPoint

Factors Associated With Metastatic Disease Testing

Results of the Cox proportional hazards models for receipt of metastatic disease testing within 18 months of treatment with curative intent among breast cancer patients are presented in Table 3. As shown in the table, a diagnosis in any of the 3 younger age groups, compared with those aged ≥ 75 years, was associated with a higher likelihood of receiving metastatic testing for patients aged < 50 years (hazard ratio [HR], 1.13), for those aged 50 years to 64 years (HR, 1.15), and for those aged 65 years to 74 years (HR, 1.13). The diagnosis of in situ (HR, 0.31) or regional (HR, 0.57) disease decreased the likelihood of receiving metastatic disease surveillance testing. White breast cancer survivors (HR, 1.13), those with a higher Charlson-Deyo comorbidity index (HR, 1.17), and those receiving care at health plan 1 (HR, 1.76) had a greater likelihood of receiving metastatic testing. We also found significant (P < .01) variation with regard to the use of metastatic testing by health plan. Table 4 shows the results of the Cox proportional hazards models for the receipt of metastatic disease testing within 18 months of treatment with curative intent among patients with colorectal cancer. Testing for metastatic disease was more likely to occur among patients aged < 50 years (HR, 1.31), or those aged 50 years to 64 years (HR, 1.25), or those aged 65 years to 74 years (HR, 1.11). Patients with more comorbidities (HR, 1.10) were also more likely to receive metastatic testing, but were less likely to do so if they were diagnosed with in situ (HR, 0.30) or local (HR, 0.60) disease. Testing for metastatic disease also varied significantly by health plan (P < .01).

Table 3. Cox Proportional Hazards Model: Metastatic Disease Testing Within 18 Months of Treatment Among Breast Cancer Survivors
Sociodemographic FactorsHR (95% CI)P
  1. Abbreviations: 95% CI, 95% confidence interval; HR, hazards ratio.

  2. Chi-square = 791.2; 11 degrees of freedom (P <.001).

 Age at diagnosis, y  
  <501.13 (1.01-1.27).03
  50-641.15 (1.05-1.26)<.01
  65-741.13 (1.03-1.25).01
  ≥75Reference
 White1.13 (1.03-1.23).01
 Median household income, $1000s1.00 (1.00-1.00).97
Disease stage  
 In situ0.31 (0.28-0.35)<.01
 Localized0.57 (0.52-0.61)<.01
 RegionalReference
Charlson-Deyo comorbidity index1.17 (1.13-1.20)<.01
Site of care  
 Health plan 11.76 (1.61-1.93)<.01
 Health plan 20.78 (0.71-0.85)<.01
 Health plan 30.98 (0.89-1.08).66
 Health plan 4Reference
Table 4. Cox Proportional Hazards Model: Metastatic Disease Testing Within 18 Months of Treatment Among Colorectal Cancer Survivors
Sociodemographic FactorsHR (95% CI)P
  1. Abbreviations: 95% CI, 95% confidence interval; HR, hazards ratio.

  2. Chi-square = 225.6; 12 degrees of freedom (P <.001).

 Age at diagnosis, y  
  <501.31 (1.06-1.63).01
  50-641.25 (1.10-1.42)<.01
  65-741.11 (0.98-1.26).10
  ≥75Reference
 Female1.08 (0.98-1.19).13
 White1.08 (0.95-1.24).25
 Median household income, $1000s1.00 (1.00-1.00).55
Disease stage  
 In situ0.30 (0.22-0.40)<.01
 Localized0.60 (0.54-0.67)<.01
 RegionalReference
Charlson-Deyo comorbidity index1.10 (1.07-1.14)<.01
Site of care  
 Health plan 11.78 (1.53-2.07)<.01
 Health plan 21.07 (0.94-1.21).29
 Health plan 30.88 (0.75-1.03).11
 Health plan 4Reference

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Among geographically diverse cohorts of breast and colorectal cancer survivors, we found deviations in the surveillance care received relative to guideline recommendations for physical examinations and local recurrence and metastatic disease screening. Although many cancer survivors appear to be receiving the minimum recommended surveillance care, a large number, regardless of age, are still not receiving minimum care recommendations whereas others are receiving nonrecommended metastatic disease testing as well as recommended care at a greater frequency than what is suggested.

In the integrated delivery systems studied in the current study, the overwhelming majority of breast and colorectal cancer survivors received the minimum recommended number of posttreatment physical examinations. In fact, the majority of both breast and colorectal cancer survivors, regardless of age or location of care, appear to receive physical examinations more frequently than is recommended by national guidelines. However, despite such examinations and national guideline recommendations, nearly one-half of the colorectal cancer cohort failed to receive a complete examination of the colon within 18 months of treatment. Conversely, the majority of breast cancer survivors received recommended recurrence testing. Regardless of cancer site, greater than 67% of survivors, particularly those who were younger, received some type of metastatic disease testing within 18 months, none of which is recommended by evidence-based guidelines. Furthermore, there were significant differences with regard to the receipt of metastatic testing among health plans that may be a result of regional and practice variations. In addition to these organizational differences, the findings of the current study highlight the patient factors associated with the use of metastatic disease testing. In fact, very little is known about what clinical factors (such as type and duration of treatment) clinicians might consider when making surveillance care recommendations to their patients. Clearly there is more to be understood about not only the multilevel factors associated with the receipt of undercare and overcare relative to guideline recommendations, but also the health and economic implications of such care use.

Compared with prior studies,10, 12, 14, 16, 18 the results of the current study appear to illustrate an increase in the receipt of recommended mammograms within 18 months of treatment with curative intent, as well as a decline in the use of nonrecommended metastatic testing among survivors of both breast and colorectal cancer. However, these improvements in care quality appeared to occur concurrently with a decrease in the receipt of recommended colon examinations among patients with colorectal cancer. Furthermore, consistent with prior studies,10, 15, 21, 24, 25 we found that age, stage, race (in patients with breast cancer only), and comorbidities at the time of diagnosis impacted the receipt of metastatic disease testing. Additional clinical evidence is needed to justify whether such variations reflect an opportunity for quality improvement from a public health perspective.

The IOM recommends that all cancer survivors receive a survivorship care plan that includes clear and effective recommendations regarding preventive practices and specific information concerning the timing and content of recommended follow-up.1 European studies have highlighted that patients exhibit a positive attitude toward follow-up care and want follow-up plans to be tailored to their specific needs.26-29 There is growing evidence that survivors in the United States have similar preferences, and that they wish to collaborate with providers in making medical decisions and want to receive care plan information directly.30-32 It remains unknown whether the patterns of follow-up care observed herein are consistent with those preferences. However, because we have no evidence that such plans were routinely used during the time of observation in any of the organizations studied herein, it is likely safe to assume that the observed patterns of follow-up are as much a result of a lack of explicit follow-up planning and care coordination as a coordinated and well-planned effort. Given that each of the participating organizations is an integrated delivery system with a comprehensive EMR system, and each routinely scores well on publicly available quality metrics,33 it appears clear that although substantive quality gains may have been made at either end of the cancer care continuum, as the IOM report points out, care delivered during the transition from cancer patient to cancer survivor more often than not is inadequate.

The conclusions of the current study should be interpreted in the light of several important limitations. First, caution should be used when generalizing study results and conclusions to other populations and settings. Study cohort members were limited to insured individuals who received their cancer care from 1 of 4 integrated health care delivery systems. Although these samples are generally similar to the general populations in their respective communities in terms of demographics, they may differ in other unmeasured respects. In particular, care should be taken in generalizing the findings of the current study to patients who are uninsured. Another limitation is that we are unable to ascertain from EMR or claims-based sources whether care received was for surveillance versus other purposes. Thus, our findings of surveillance care overuse may be partially explained by our lack of information regarding symptoms and laboratory findings and our inability to differentiate between procedures received for screening versus diagnostic purposes. However, a prior study found that claims data capturing procedures and visit use for characterizing guideline adherence was comparable to documentation found in the medical record, and that administrative data could be used to describe patterns of follow-up care.13 A further limitation stems from grouping patients, with heterogeneity in prognoses, into general disease stages. Patients within these broad groups and the clinicians treating them may resort to different surveillance practices based on perceived differences in prognosis. Yet it is important to note that current guidelines do not differ in their recommendations for surveillance care based on these or other factors.

Compared with other phases of cancer control and prevention, surveillance care among cancer survivors appears to be understudied. The findings of the current study highlight the wide variations that exist in cancer surveillance care among seemingly clinically similar patients and across different age groups. This variation combined with the overall lack of concordance with established clinical practice guidelines highlights the need for research exploring both whether observed variations are driven by patient preferences and reflect informed decision-making and how survivorship planning as outlined by the IOM can impact such variations. Furthermore, given the survival advantage for patients aged < 65 years, it is important to consider the care trajectory and its implications among this subset of cancer patients.

FUNDING SUPPORT

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES

Funding for this research was provided by National Cancer Institute (NCI) grants R01 CA114204, R01 CA114204-03S1, and R25 CA116339 and NCI Cooperative Agreement U19 CA79689. The following staff members provided data processing and analysis support for this study: Erin Keast, Erin Masterson, Donald Bachman, and Arthur Dixon (Kaiser Permanente Northwest, Portland, OR); Christi Hanson and Arvind Ramaprasan (Group Health Cooperative, Seattle, WA); Stephanie Latimer and Gwyn Saylor (Kaiser Permanente Colorado, Denver, CO); and Nonna Akkerman and Elizabeth Dobie (Henry Ford Health System, Detroit, MI).

CONFLICT OF INTEREST DISCLOSURES

The authors made no disclosures.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. MATERIALS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. FUNDING SUPPORT
  8. REFERENCES