Table 1.

Study Characteristics

Author	Year	Country	Sample characteristics	Analysis	Variables/Measures	Outcome	Quality appraisal
Andersen and Urban [36]	1998	USA	Breast cancer survivors n = 485 50–80 years old 3-20+ years post-diagnosis	Multiple logistic regression	Receipt of mammogram, usual source of care,1 recommendation by physician for mammogram and insurance coverage	Receipt of mammogram	Average
Andrykowski and Burris [45]	2010	USA	SEER database Breast cancer survivors n = 42 Colorectal cancer survivors n = 33 Hematological cancer survivors n = 38 1–5 years post-diagnosis Aged 25–75 years old	Multiple regression	Socio-demographics, cancer characteristics, mental health resource questionnaire	Use of formal and informal mental health services	Very good
Boehmer et al. [34]	2010	USA	Colorectal cancer survivors Aged 22–92 years old n = 253	Cox proportional hazard models	Colonoscopies, sigmoidoscopy, cancer type, stage, co-morbidities, outpatient visits, socio-demographics	Receipt of colorectal surveillance procedures	Very good
Cooper et al. [29]	2000	USA	SEER-MEDICARE database Colorectal cancer survivors Localised disease Surgically treated >65 years old n = 5, 716	Chi-square test	Socio-demographics, inpatient claims, outpatient claims, use of endoscopic procedures (colonoscopy, polypectomy or biopsy)	Receipt of colorectal surveillance procedures	Very good
Cooper and Payes [28]	2006	USA	SEER-MEDICARE database Colorectal cancer survivors >65 years old n = 62, 882 survived 1 year follow-up n = 35, 784 survived 3 year follow-up	Logistic regression	Medicare claims2 for colonoscopy, sigmoidoscopy or barium enema, co-morbidities	Use of surveillance procedures for colorectal cancer within 3 years of diagnosis	Very good
Cooper, Kou and Reynolds [31]	2008	USA	SEER database Colorectal cancer survivors >65 years old n = 9, 426	Multivariate regression	Number of physician visits, receipt of carcino-embryonic antigen blood test (CEA),3 colonoscopy, CT and PET scans	Adherence to guidelines for cancer follow-up	Good
Doubeni et al. [27]	2006	USA	Breast cancer survivors n = 797 at baseline (end of treatment) n = 262 after 5 yrs >55 years old 4 geographically diverse Health Maintenance Organisations (HMOs).4	Generalised estimated equations (GEE)	Receipt of mammograms. age, date and stage at/of diagnosis, treatment. co-morbidities. visits to primary care provider (primary care physician) and outpatient visits	Receipt of yearly mammogram and visits to physicians	Very good
Earle et al. [23]	2003	USA	SEER database Breast cancer survivors > 65 years old, n = 5,965 Controls n = 6,062	Multivariate regression	Frequency of visits to primary care physician, oncologists, other and teaching hospitals, receipt of flu vaccine, lipid test, cervical exam, colon exam, bone densitometry and diabetes test	Visits to physicians and receipt of preventive medicine	Very good
Earle and Neville [19]	2004	USA	SEER database Colorectal cancer survivors > 65 years old n = 14,884	Logistic regression	Co-morbidities, socio-demographics, receipt of flu vaccine, lipid testing, bone densitometry and cervical screening	Visits to physicians and receipt of preventive medicine	Very good
Earle, Neville and Fletcher [43]	2007	USA	Breast, lymphoma, colorectal, melanoma and other cancer survivors Mean age 60 years n = 1,111 Controls n = 4,444	Logistic regression `	Mental health diagnoses, co-morbidities, socio-demographics, use of primary care physician, oncologist, psychiatrists, psychologists, social workers and inpatient hospitalisations (both general and mental).	Use of mental health provider services	Good
Ellison et al. [33]	2003	USA	SEER database Colorectal cancer survivors >65 years old n = 52, 105	Kaplan-Meier survival analysis Unconditional regression analysis Cox regression	Socio-demographic, hospital and clinical characteristics, receipt of colonoscopy, sigmoidoscopy, endoscopy and barium enema	Differential receipt of colonoscopy, sigmoidoscopy, endoscopy and barium enema by race	Good
Gray et al. [41]	2000	Canada	Breast cancer survivors n = 731 Histologically confirmed and invasive	Stepwise logistic regression	Use of specialised supportive care services, wish to use services that were not accessed, social and demographic characteristics.	Use of professional supportive care services provided by the Ontario health care system	Very good
Gray et al. [42]	2002	Canada	Breast cancer survivors 63 % <60 years old 23–36 months post-diagnosis n = 731	Logistic regression	Supportive care from physicians and nurses, socio-demographics, illness and treatment information	Use of professional supportive care	Good
Grunfeld et al. [16]	1999	UK	Breast cancer survivors n = 148 Two district general hospitals	Two-tailed t-test and chi-square	Record of visits, average cost of visits, out-of patient expenses, waiting times, lost earnings and lost earnings of accompanying person	GP follow-up vs. Hospital follow-up. Cost-effectiveness and cost to patient,	Average
Grunfeld et al. [17]	2011	Canada	Breast cancer survivors n = 408 Nine tertiary cancer centres	Two-tailed t-test	Use of survivorship care plans (vs. no survivorship care plans) in primary care physician led follow-up. Frequency of visits to oncologists.	Primary care physician led follow-up	Very good
Keating et al. [25]	2006	USA	SEER-MEDICARE database Breast cancer survivors Stage 1 or 2 Underwent surgery >65 years old	Repeated-measures logistic regression	Mammogram receipt, visits to primary care physician medical oncologist, general surgeon, radiation oncologist and other specialists, socio-demographics	Factors related to mammography use	Very good
Keating et al. [11]	2007	USA	SEER database Breast cancer survivors >65 years old n = 37,967 in year 1 n = 30,406 in year 2 n = 23,016 in year 3	Repeated-measures logistic regression	Receipt of bone scans, tumour antigen tests (TAT), Chest x-rays and other abdominal/chest imaging, frequency of visits to physicians and socio-demographics	Receipt of a number of surveillance procedures and visits to physicians over time	Very good
Khan et al. [38]	2010	UK	GPRD database Breast cancer survivors N = 18, 612 Colorectal cancer survivors N = 5, 764 Prostate cancer survivors N = 4, 868 >30 years old 5 years post-diagnosis Controls N = 116,418	Multivariate regression	Socio-demographics, use of primary care, frequency of visits	Primary care consultations	Very good
Khan, Watson and Rose [20]	2011	UK	GPRD database Prostate cancer survivors N = 4,868 Breast cancer survivors N = 18,612 Colorectal cancer survivors N = 5,764 Controls N = 145,662	Logistic regression	Co-morbidities, screening (PSA, cervical, mammogram), receipt of preventative procedures and socio-demographics	Receipt of screening and preventative care	Very good
Knopf et al. [37]	2001	USA	SEER database Colorectal cancer survivors >65 years old n = 52, 283	Kaplan-Meier survival analysis	Receipt of colonoscopy, sigmoidoscopy, endoscopy and barium enema, age, tumour stage at diagnosis and year of diagnosis	Receipt of bowel surveillance procedures	Very Good
Lafata et al. [30]	2001	USA	Colorectal cancer survivors n = 251	Kaplan-Meier survival analysis Cox proportional hazards	Socio-demographics, receipt of colonoscopy, CEA, barium enema, chest x-ray, MRI’s, ultrasounds and liver analysis	Receipt of colon screening procedures and other procedures	Very good
Mahboubi et al. [15]	2007	France	Colorectal cancer survivors <65 years old N = 389	Logistic regression	Co-morbidities, chest radiograph, abdominal ultrasound, colonoscopy, CT, TAT, blood tests and reason for testing (routine or symptomatic)	Characteristics associated with visits to GPs	Very good
Mandelblatt et al. [13]	2006	USA	Breast cancer survivors n = 418 Stage 1 and 2	Multivariate linear regression	Calendar diary of health service use, socio-demographics, cancer treatment information, co-morbidities and psychological status survey	Patterns and determinants of health service use	Very good
Mayer et al. [35]	2007	USA	NCI 2003 HINTS5 n = 619 Breast cancer survivors n = 119 Prostate cancer survivors n = 62 Colorectal cancer survivors n = 49 Others n = 389	Logistic regression	Based on the health belief model (HBM),6 cancer communication, cancer history, general cancer knowledge, cancer risk and screening, health status and demographics.	Screening practices and beliefs	Very good
McBean, Yu and Virnig [39]	2008	USA	SEER database: Uterine cancer survivors >65 years old n = 14,575 Controls n = 58,420	Multivariate logistic regression Generalised equation modelling	Receipt of flu vaccine, bone densitometry, colorectal screening and mammogram no. of physician services and socio-demographics	Use of preventive services and frequency of physician visits	Very good
Mols, Helfenrath and van de Poll-Fanse [14]	2007a	Netherlands	Endometrial cancer Prostate cancer Non-Hodgkin’s lymphoma survivors n = 1,112	Linear regression Multivariate linear regression	SF-36, self-reported health service use, frequency of visits, co-morbidities and socio-demographics	Patterns of physician use	Very good
Mols, Coebergh and van de Poll-Fanse [22]	2007b	Netherlands	Endometrial cancer Prostate cancer, Hodgkin’s and non-Hodgkin’s lymphoma survivors n = 1,231	Chi-square and multivariate logistic regression	Co-morbidity, socio-demographics, use of medical specialist, general practitioner, additional services (physiotherapist. and psychologist)	Frequency of physician use	Very good
Oleske et al. [47]	2004	USA	Breast cancer survivors Aged between 21–65 years n = 123	Multivariate logistic regression	Use and frequency of physician and admissions, services in past 12 months. reasons for hospitalisations, SRS (social responsiveness scale) and CES-D (depression scale)	Determination of factors associated with hospitalisation	Very good
Peuckmann et al. [12]	2009	Denmark	Breast cancer survivors n = 1,316 Controls n = 4,865	Risk ratios and multiple logistic regression analysis	Frequency of physical visits, socio-demographics, physical activity and BMI. HR-QOL (SF-36) and chronic pain	Frequency and determinants of health service use	Very good
Schapira, McAuliffe and Nattinger [32]	2000	USA	SEER database Breast cancer survivors >65 years old n = 3,885	Logistic model	Receipt of mammogram, co-morbidity, socio-economic status (SES) and preventive treatment received	Receipt of Mammogram over two year period	Good
Schootman et al. [44]	2008	USA	SEER database Breast cancer survivors >65 years old n = 47, 643	Restricted iterative generalised least squares and first-order marginal quasi-likelihood estimation analysis	Frequency of Ambulatory-Care-Sensitive Hospitalizations (ACSH)7 SES, co-morbidity, demographics, availability of medical care, visits to primary care physician and oncologists	Frequency of Ambulatory-Care-Sensitive Hospitalizations	Very good
Simpson, Carlson and Trew [18]	2001	USA	Breast cancer survivors Time point 1 n = 46 Time point 4 n = 30 Controls Time point 1 n = 43 Time point 4 n = 25	ANOVA	Average cost of care, no. of cancer centre visits and a number of psychological distress indicators including BDI, POMS and Mental adjustment to cancer scale	Billing of Health care as a proxy to use. Visits to cancer centre Correlation of billing to distress.	Good
Snyder et al. [9]	2008a	USA	SEER database Colorectal cancer survivors >65 years old n = 1,541	Poisson regression and logistic regression	Clinical and socio-demographic characteristics, visits to primary care physician, oncologist or other physicians. Receipt of influenza vaccine, cholesterol screening, mammogram, cervical screening and bone densitometry	Frequency of physician visits and receipt of preventive care	Very good
Snyder et al. [10]	2008b	USA	SEER database Colorectal cancer survivors >65 years old n = 20,068	Poisson regression and logistic regression analysis	Co-morbidities, socio-demographics, visits to primary care physician, oncologist and other physicians, receipt of influenza vaccine, cholesterol screening, mammogram, and bone densitometry	Visits to physicians and receipt of preventive care	Good
Snyder et al. [24]	2009a	USA	SEER database Breast cancer survivors >65 years old n = 23, 73 Controls n = 23, 731	Poisson regression and logistic regression analysis	Use of physician and oncology services, receipt of 5 preventive care services and socio-demographics.	Visits to physicians and oncologists and preventive medicine	Good
Snyder et al. [26]	2009b	USA	SEER database Breast cancer survivors >65 years old Stages 1–3 n = 1,961 Controls n = 1,961	Poisson regression and logistic regression analysis	Co-morbidities, clinical and demographic characteristics, visits to primary care physician, oncologists and other physicians	Frequency of visits to physicians	Good
Van de Poll-Fanse et al. [21]	2006	Netherlands	Breast cancer survivors Invasive n = 183	Logistic regression	Co-morbidities, spontaneously reported problems, use of GP, medical specialist and physiotherapist, health status and psychological well-being	Use of physician services	Good
Yu, McBean and Virnig [40]	2007	USA	SEER database Colorectal cancer survivors >65 years old n = 112, 737.	Logistic regression and poisson regression	Socio-demographic characteristics, co-morbidities, receipt of mammogram, visits to primary care physician, Gynaecologists only, oncologists and other	Receipt of mammogram and visits to physicians	Good

¹Usual source of care refers to whether an individual receives care from the same physician or different physicians; ²Medicare is a government-funded medical care plan in USA, whereby individuals aged 65 and over that covers medical expenses such as doctor's visits, hospital stays, drugs and other treatment; ³CEA testing is used as a tumour marker for particular cancers, such as colorectal; ⁴HMOs provide their members with medical services for a fixed fee; ⁵NCI HINTS is the Health Information National Trends Survey, which collects nationally represented information on how the American public find and use information on cancer; ⁶Developed by Hochbaum (1958) is an explanatory and predictive model of health behaviours and includes attitudes and beliefs of an individual; ⁷ACSH are hospitalizations which could have been prevented if primary care services had been initially accessed by the individual.