This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Snyder, C. F. Articles by Frick, K. D. Search for Related Content PubMed PubMed Citation Articles by Snyder, C. F. Articles by Frick, K. D. Social Bookmarking                            What's this?

Preventive Care for Colorectal Cancer Survivors: A 5-Year Longitudinal Study

Claire F. Snyder, Craig C. Earle, Robert J. Herbert, Bridget A. Neville, Amanda L. Blackford, Kevin D. Frick

From the Division of General Internal Medicine, Department of Medicine; Department of Oncology, Johns Hopkins School of Medicine; Department of Health Policy & Management, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD; and Division of Population Sciences, Department of Medical Oncology, Dana-Farber Cancer Institute, Boston, MA

Corresponding author: Claire Snyder, PhD, Johns Hopkins School of Medicine, 624 N Broadway, Room 657, Baltimore, MD 21205; e-mail: csnyder{at}jhsph.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
Purpose To explore the mix of physician specialties that long-term survivors visit and how the mix relates to preventive care.

Patients and Methods Using the Surveillance, Epidemiology, and End Results–Medicare database, we conducted a retrospective, longitudinal study of stage I to III Medicare fee-for-service colorectal cancer patients diagnosed in 1997. We examined physician visits and preventive care each year for 5 years, starting 366 days postdiagnosis, and how preventive service receipt related to the physician mix seen: primary care provider (PCP) only, oncologist only, both, or neither.

Results A total of 1,541 patients met the eligibility criteria (mean age, 76; 43% male; 85% white). During 5 years, PCP visits increased from a mean of 4.2 to 4.7 (P < .0001), and oncology visits decreased from 1.3 to 0.5 (P < .0001). Survivor care by PCPs only increased from 44% to 62%, whereas shared care by PCPs and oncologists dropped from 37% to 21% (P < .0001). Survivors who saw both PCPs and oncologists were most likely to receive influenza vaccination, mammograms, and cervical cancer screening; survivors who saw PCPs only were most likely to receive cholesterol screening and bone densitometry. Higher socioeconomic status was associated with increased influenza vaccination, mammograms, and cervical cancer screening (P < .05). Over time, there was a decrease in mammography and cervical cancer screening and an increase in influenza vaccination (P < .05).

Conclusion As oncologists become less involved in survivor care, cancer-related screening decreases significantly. These results support the need for survivorship care plans that explicitly outline the roles of PCPs and oncologists in sharing care for cancer survivors, and how these roles may change over time.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
The Institute of Medicine report, From Cancer Patient to Cancer Survivor: Lost in Transition, notes that quality follow-up care is essential for cancer survivors.¹ Of the 10 million cancer survivors in the United States, approximately 1 million survived colorectal cancer,¹ resulting from its frequent incidence (the third most common cancer) and high survival rates (84% at 1 year, 64% at 5 years).² Approximately 67% of colorectal cancer cases occur in adults older than age 65, making it common among the Medicare population.³

Little is known about the appropriate primary and cancer-related care of these survivors as they transition from active treatment to survivorship—or who should be responsible for providing that care. Treatment guidelines recommend survivors of stage II to III colorectal cancer get a history and physical every 3 to 6 months during the first 5 years of survivorship.^4-6 However, Lafata⁷ found that only approximately 75% of colorectal cancer survivors had two surveillance visits within 18 months.

The guidelines described above focus on surveillance for cancer recurrence. They do not address cancer survivors' needs for primary and preventive care, nor do they specifically address what physician specialty survivors should see for their surveillance. A recent study in breast cancer found that primary care providers (PCPs) are to monitor survivors for recurrence-related serious clinical events as well as cancer specialists.⁸ A 1996 survey of medical oncologists found that they are often a patient's primary caregiver during active treatment, but that they spend little time providing primary care and, given the choice, they would prefer to increase their cancer care rather than primary care.⁹

Earle and Neville¹⁰ compared 5-year colorectal cancer survivors to matched noncancer controls to examine patterns of care, and found that cancer survivors were less likely to receive recommended care for chronic conditions. Cancer survivors monitored by both oncology specialists and PCPs were more likely to receive appropriate care, with PCPs more likely to provide noncancer-related recommended care and oncology specialists more likely to provide cancer-related recommended care. A study of breast cancer survivors produced similar findings.¹¹

These studies provide only a cross-sectional picture of the care received by cancer survivors 5 years post-treatment; they do not observe survivors longitudinally to examine how the mix of physicians visited and care received change over time. We used longitudinal data to examine physician visits and preventive care for long-term cancer survivors during a 5-year period. We chose long-term survivors for this longitudinal study because their care provides the clearest picture of how these transitions occur over time.

	PATIENTS AND METHODS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
Research Design
This retrospective, longitudinal cohort study examined the patterns of physician visits and preventive care of colorectal cancer survivors each year for 5 years after the end of active treatment. For each year, we calculated the number of visits to different physician types and the percentage of survivors who received preventive care services. Receipt of preventive care was then compared based on the physician mix seen: both PCPs and oncology specialists, PCPs only, oncology specialists only, and neither PCPs nor oncology specialists.

Data Source
We used the Surveillance, Epidemiology, and End Results (SEER) –Medicare database, which links the data from the SEER cancer registry program to Medicare claims.¹² During 1997, there were 13 SEER registries, covering a representative sample of approximately 14% of the US population.¹³ In the SEER-Medicare database, these SEER data are complemented by the Medicare claims data, which provide longitudinal information on the Medicare-covered health care services received. For this study, the Medicare claims data were used to determine the number of health care visits and to what physician specialties, and receipt of covered preventive services. Currently, Medicare-claims data are available through 2003; thus, the 1997 diagnostic cohort includes the most recent data available for follow-up through 5 years of survival (1998 to 2003).

Study Participants
Survivors were eligible for this study if they met the following criteria: (1) diagnosis of stage I to III colorectal cancer in 1997; (2) residing in a SEER region at diagnosis; (3) eligible for Medicare at diagnosis based on age, and enrolled continuously in the fee-for-service program (Parts A and B) during the period of analysis (1998 to 2003); (4) survived through 2003 with no subsequent malignant diagnoses; (5) did not receive chemotherapy or radiation during the survivorship period (defined as beginning 366 days after diagnosis); and (6) did not enroll in hospice. These eligibility criteria were used to reduce the possibility that the patterns of physician visits and preventive care received would be affected by long-term sequelae of the cancer, its treatment, or any recurrence or subsequent diagnoses.¹⁰

Variables
Data from the first year of survivorship were used to examine the age, sex, race, stage of disease, SEER registry site, and comorbidity of the population. The comorbidity index was calculated based on the claims data using the Charlson score¹⁴ as implemented by Deyo¹⁵ and modified by Klabunde.¹⁶

Each survivorship year was calculated using a 365-day period, with the first year of survivorship starting on day 366 after the SEER date of diagnosis. To identify the mix of physician specialists seen each year, we used the physician specialty reported to Medicare. PCPs included general practice, internal medicine, family practice, obstetrics/gynecology, geriatrics, and multispecialty group practice. Oncology specialties included medical oncology, hematology-oncology, general surgery, colorectal surgery, surgical oncology, and radiation oncology.

The preventive measures used were based on those used by Earle and Neville¹⁰ and included influenza vaccination, cholesterol screening, mammography for women younger than age 76, cervical cancer screening, and bone densitometry. Each of these measures was assessed each year during the 5-year period. Indicators for colorectal cancer screening and follow-up were not included because this analysis focuses on preventive care services, unrelated to the cancer diagnosis, rather than surveillance for recurrence.

Analyses
Cohort characteristics were described using the variables outlined above. The mean and standard deviation values of physician visits to PCPs, oncology specialists, and others were calculated each year for the 5-year survivorship period overall and by stage of disease. Poisson regression with generalized estimating equations was used to model the number of visits as a function of year to test for changes in the frequency of visits over time. The percentage of survivors receiving each preventive service each year for the 5-year period was then calculated.

Based on the provider mix seen, survivors were classified into four groups: (1) visits to a PCP but not an oncology specialist, (2) visits to an oncology specialist but not a PCP, (3) visits to both a PCP and oncology specialist, and (4) visits to neither a PCP nor an oncology specialist. Survivors who had no physician visits or only "other" physician visits were included in the latter category. Survivors were categorized separately for each year of follow-up, and the percentages of survivors in each category each year were examined. Thus, survivors could change categories depending on the physician mix they saw each year. A repeated-measures analysis was used to test for marginal homogeneity in the categorization of survivors over time. Logistic regression models with generalized estimating equations were also used to describe the characteristics associated with seeing both provider types compared with the other three categories. We investigated age at diagnosis (65 to 74, 75 to 84, 85 years), race (white v other), comorbidity index (0, 1, 2+), sex, socioeconomic status (quintiles as a continuous variable), stage, and urban/rural location, while controlling for year of follow-up and SEER registry site.

Finally, we examined the effect of provider mix on the receipt of preventive care, both unadjusted and controlling for total number of visits per year (zero to four, five to seven, eight to 12, 13+), age at diagnosis, race, stage, comorbidity index, socioeconomic status, sex, urban/rural location, year of follow-up, and SEER registry site.

	RESULTS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
A total of 1,541 survivors met the eligibility criteria. The final survival population was identified as follows: 13,340 patients, residing in a SEER region, were diagnosed with colorectal cancer in 1997; of those, 8,915 were diagnosed with stage I to III disease; of those, 2,222 were eligible for Medicare based on age and continuously enrolled onto the fee-for-service program between 1998 and 2003; of those, 1,817 patients survived through 2003 and had no subsequent malignant diagnoses; of those survivors, 1,556 did not receive chemotherapy or radiation during the survivorship period; and of those, 1,541 did not enroll in hospice. Among the 1,541 survivors meeting the eligibility criteria, the mean age was 76.0 years, 43% were male, and 85% were white (Table 1).

In unadjusted ² analyses combining data across the 5 years, survivors who saw both a PCP and oncology specialist were most likely to receive each preventive service listed in Table 4 (P < .0001). In analyses controlling for total number of visits, age at diagnosis, sex, race, stage, comorbidity index, socioeconomic status, urban/rural location, year of follow-up, and SEER registry site, the mix of physician specialties seen remained statistically significant for all five preventive services (Table 5). Although the overall variance explained by the model as estimated by pseudo-R² was low (ranging from 0.05 to 0.09), adding provider mix to the model resulted in a percent increase in pseudo-R² ranging from 8% to 58%. Survivors seeing both a PCP and oncology specialist were most likely to receive influenza vaccination, mammograms, and cervical cancer screening. Survivors who saw a PCP only were most likely to receive cholesterol screening and bone densitometry. There were no statistically significant differences in care between survivors who saw a medical oncologist only compared with those who saw a general surgeon only (though bone densitometry had insufficient sample sizes for estimation).

View larger version (16K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 1. Percent-of-service use by comorbidity index level (0, 1, 2+). Percent-of-service use adjusted for provider mix; total number of visits; age at diagnosis; sex; race; stage; socioeconomic status; urban/rural location; year of follow-up; and Surveillance, Epidemiology, and End Results (SEER) registry site (mammography indicator not limited to women younger than 76 years of age in this analysis).

	DISCUSSION

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

Although we cannot make definitive conclusions based on these observational data, it is noteworthy that cancer screening decreases as oncology specialists become less involved in cancer survivors' care. That survivors who saw both a PCP and oncology specialist were most likely to receive influenza vaccination, mammography, and cervical cancer screening, whereas patients who saw a PCP only were most likely to have cholesterol screening and bone densitometry, suggests a possible division of labor between oncology specialists and PCPs. As oncology specialists become less involved in survivors' care, the system may fail to identify the need to transfer responsibility for these preventive services to the PCP.

Another possible explanation for these observed changes is the controversy regarding the usefulness of screening in the elderly or those with comorbidities. There was no clear relationship between the number of comorbidities and receipt of preventive services, suggesting either that the comorbidity measure is insensitive or that some service use is inappropriate. Older age was statistically significantly related to decreases in cholesterol screening, cervical cancer screening, and bone densitometry. The appropriate frequency of cervical cancer screening and age at which screening should stop is particularly controversial.¹⁷ Because we only assessed mammography in survivors younger than age 76 (based on the quality indicator used by Earle¹⁰), age was not a factor. We also analyzed the mammography indicator without an age limit and found that younger women were statistically significantly more likely to get a mammogram. Clinicians may be making judgments regarding survivors' life expectancy in determining who should receive preventive care. However, a previous analysis demonstrated that cancer survivors are less likely to receive preventive care compared with matched noncancer controls.¹⁰ Whether this disparity results from clinical judgments of life expectancy or deficiencies in care for cancer survivors requires further research. That increases in socioeconomic status were associated with increases in preventive care also suggests a possible disparity.

Compared with the rates of preventive service use among 5-year colorectal cancer survivors,¹⁰ the preventive service rates at 5 years were generally lower in this study: cholesterol screening, 36.5% v 31.0%; mammography, 54.0% v 48.4%; and cervical cancer screening, 17.8% v 11.0%. Influenza vaccinations were slightly higher in our sample (54.3% v 53.2%), and bone densitometry was also higher (10.1% v 4.2%). The rates of preventive care for the matched noncancer controls in the study by Earle and Neville¹⁰ were influenza vaccination, 55.4%; cholesterol screening, 39.4%; mammography, 51.5%; cervical cancer screening, 21.9%; and bone densitometry, 5.7%. Thus, the rates of preventive service receipt in our sample tended to be lower than both the cancer survivors and the noncancer controls reported by Earle and Neville.¹⁰

One reason for the difference may be the time window used by Earle and Neville¹⁰ compared with our analysis. Because we observed longitudinally for 5 years, each year was defined by a 365-day window. In contrast, Earle and Neville¹⁰ looked at whether the preventive services had occurred at any point during a 2-year period and did not take into account whether annual services were performed once or twice during that period. Because we observed survivors for multiple years, we could not expand our yearly time window longer than 365 days without risking double counting. In practice, annual events might not occur every 12 months exactly but might occur a few months before or after.

Because this is a secondary data analysis using administrative data, we have no information regarding why a service was not delivered—whether the provider failed to offer it or whether the patient refused. Furthermore, our data did not include preventive care the survivors may have received before their diagnosis or the therapies they received during their active cancer treatment. Because prediagnosis-service use may be one of the strongest predictors of survivorship service use, it is difficult to discern whether the findings related to provider mix are real or reflect baseline use or other factors.

This study had several other limitations. First, as noted previously, the SEER-Medicare database only includes fee-for-service Medicare patients; therefore, we do not know how Medicare cancer survivors in the managed care setting may differ. Second, we do not have data on the physicians (eg, age, sex) or on the relative costs of visits to PCPs versus oncology specialists. Third, because we used administrative claims data, we would not have captured preventive services that were provided but not billed for. Fourth, there may be some misclassification of provider type if the provider type on file with Medicare is not reflective of a physician's actual practice.

Finally, individuals had to survive without subsequent malignant diagnosis for 6 years postdiagnosis to be included in our sample. This eligibility criterion probably explains the high percentage of patients in our sample with early-stage disease. An alternative approach would have included individuals until subsequent malignant diagnosis or death. Our approach has the benefit of observing the same cohort of survivors all 5 years without the potential of confounding due to death or subsequent diagnosis. Specifically, patients who died or had a subsequent malignant diagnosis may have had a significant number of physician visits, which would have skewed our findings, and these patients would not have been good candidates for preventive care. Although the findings of this study were derived from long-term survivors, we believe the interpretation of the results would apply equally to patients who are not long-term survivors.

Our results support the need for survivorship care plans, as recommended by the Institute of Medicine.¹ These survivorship care plans and treatment guidelines should define the roles and responsibilities of PCPs and oncology specialists, and take into account that, over time, survivors may visit oncology specialists less often and rely solely on PCPs more often. Developing survivorship care plans with clearly delineated roles for PCPs and oncology specialists can help ensure that quality care is provided in a coordinated and efficient manner.

	AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
The author(s) indicated no potential conflicts of interest.

	AUTHOR CONTRIBUTIONS

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
Conception and design: Claire F. Snyder, Craig C. Earle, Kevin D. Frick

Financial support: Kevin D. Frick

Collection and assembly of data: Robert J. Herbert, Bridget A. Neville

Data analysis and interpretation: Claire F. Snyder, Craig C. Earle, Robert J. Herbert, Bridget A. Neville, Amanda L. Blackford, Kevin D. Frick

Manuscript writing: Claire F. Snyder, Craig C. Earle, Bridget A. Neville, Amanda L. Blackford, Kevin D. Frick

Final approval of manuscript: Claire F. Snyder, Craig C. Earle, Robert J. Herbert, Bridget A. Neville, Amanda L. Blackford, Kevin D. Frick

	NOTES

 
Supported by an unrestricted research grant from Pfizer Inc.

Presented at the 43rd Annual Meeting of the American Society of Clinical Oncology, June 1-5, 2007, Chicago, IL.

Authors' disclosures of potential conflicts of interest and author contributions are found at the end of this article.

	REFERENCES

TOP ABSTRACT INTRODUCTION PATIENTS AND METHODS RESULTS DISCUSSION AUTHORS' DISCLOSURES OF... AUTHOR CONTRIBUTIONS REFERENCES

 
1. Institute of Medicine: From Cancer Patient to Cancer Survivor: Lost in Transition. Washington, DC, National Academy Press, 2005

2. American Cancer Society: Cancer Facts & Figures 2007. Atlanta, GA, American Cancer Society, 2007

3. National Cancer Institute: Cancer stat fact sheets: Cancer of the colon and rectum. http://seer.cancer.gov/statfacts/html/colorect.html

4. Desch CE, Benson AB III, Somerfield MR, et al: Colorectal cancer surveillance: 2005 update of an American Society of Clinical Oncology practice guideline. J Clin Oncol 23:8512-8519, 2005[Abstract/Free Full Text]

5. National Comprehensive Cancer Network: NCCN clinical practice guidelines in oncology: Colon Cancer. V. 2. 2006. http://www.nccn.org/professionals/physician_gls/PDF/colon.pdf

6. National Comprehensive Cancer Network: NCCN clinical practice guidelines in oncology: Rectal cancer. V. 2. 2006. http://www.nccn.org/professionals/physician_gls/PDF/rectal.pdf

7. Lafata JE, Simpkins J, Schultz L, et al: Routine surveillance care after cancer treatment with curative intent. Med Care 43:592-599, 2005[CrossRef][Medline]

8. Grunfeld E, Levine MN, Julian JA, et al: Randomized trial of long-term follow-up for early-stage breast cancer: A comparison of family physician versus specialist care. J Clin Oncol 24:848-855, 2006[Abstract/Free Full Text]

9. American Society of Clinical Oncology: Status of the medical oncology workforce. J Clin Oncol 14:2612-2621, 1996[Abstract]

10. Earle CC, Neville BA: Under use of necessary care among cancer survivors. Cancer 101:1712-1719, 2004[CrossRef][Medline]

11. Earle CC, Burstein HJ, Winer EP, et al: Quality of non-breast cancer health maintenance among elderly breast cancer survivors. J Clin Oncol 21:1447-1451, 2003[Abstract/Free Full Text]

12. National Cancer Institute: SEER-Medicare: Brief description of the SEER-Medicare database. http://healthservices.cancer.gov/seermedicare/overview/brief.html

13. National Cancer Institute: Number of persons by race and Hispanic ethnicity for SEER participants. http://seer.cancer.gov/registries/data.html

14. Charlson ME, Pompei P, Ales KL, et al: A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chron Dis 40:373-383, 1987[CrossRef][Medline]

15. Deyo RA, Cherkin DC, Ciol MA: Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol 45:613-619, 1992[CrossRef][Medline]

16. Klabunde CN, Potosky AL, Legler JM, et al: Development of a comorbidity index using physician claims data. J Clin Epidemiol 53:1258-1267, 2000[CrossRef][Medline]

17. Clinical Practice Committee of the American Geriatric Society: Position paper on screening for cervical carcinoma in older women, 2000. http://www.americangeriatrics.org/products/positionpapers/cer_carc_2000.shtml

Submitted March 30, 2007; accepted November 8, 2007.

CiteULike    Complore    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?

This article has been cited by other articles:

				C. Rabin Promoting Lifestyle Change Among Cancer Survivors: When Is the Teachable Moment? American Journal of Lifestyle Medicine, September 1, 2009; 3(5): 369 - 378. [Abstract] [PDF]

				W. Y. Cheung, B. A. Neville, D. B. Cameron, E. F. Cook, and C. C. Earle Comparisons of Patient and Physician Expectations for Cancer Survivorship Care J. Clin. Oncol., May 20, 2009; 27(15): 2489 - 2495. [Abstract] [Full Text] [PDF]

				L. Nekhlyudov "Doc, Should I See You or My Oncologist?" A Primary Care Perspective on Opportunities and Challenges in Providing Comprehensive Care for Cancer Survivors J. Clin. Oncol., May 20, 2009; 27(15): 2424 - 2426. [Full Text] [PDF]

				P. C. Nathan, J. S. Ford, T. O. Henderson, M. M. Hudson, K. M. Emmons, J. N. Casillas, E. A. Lown, K. K. Ness, and K. C. Oeffinger Health Behaviors, Medical Care, and Interventions to Promote Healthy Living in the Childhood Cancer Survivor Study Cohort J. Clin. Oncol., May 10, 2009; 27(14): 2363 - 2373. [Abstract] [Full Text] [PDF]

				L. N. Shulman, L. A. Jacobs, S. Greenfield, B. Jones, M. S. McCabe, K. Syrjala, L. Diller, C. L. Shapiro, A. C. Marcus, M. Campbell, et al. Cancer Care and Cancer Survivorship Care in the United States: Will We Be Able to Care for These Patients in the Future? J. Oncol. Pract, May 1, 2009; 5(3): 119 - 123. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Services Email this article to a colleague Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Save to my personal folders Download to citation manager Rights & Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Snyder, C. F. Articles by Frick, K. D. Search for Related Content PubMed PubMed Citation Articles by Snyder, C. F. Articles by Frick, K. D. Social Bookmarking                            What's this?