Educational outcomes among survivors of childhood cancer in British Columbia, Canada

Report of the Childhood/Adolescent/Young Adult Cancer Survivors (CAYACS) Program

Authors

  • Maria Lorenzi MSc,

    1. Cancer Control Research Program, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Amy J. McMillan MA,

    1. Cancer Control Research Program, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    2. Department of Educational and Counseling Psychology and Special Education, University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Linda S. Siegel PhD,

    1. Department of Educational and Counseling Psychology and Special Education, University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Bruno D. Zumbo PhD,

    1. Department of Educational and Counseling Psychology and Special Education, University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Victor Glickman PhD,

    1. Faculty of Education, University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • John J. Spinelli PhD,

    1. Cancer Control Research Program, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    2. Department of Health Care and Epidemiology, University of British Columbia, Vancouver, British Columbia, Canada
    3. Department of Statistics and Actuarial Science, Simon Fraser University, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Karen J. Goddard MB, ChB,

    1. Department of Radiation Oncology, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Sheila L. Pritchard BM, BS,

    1. Division of Oncology, Hematology, and Bone Marrow Transplant, British Columbia Children's Hospital and University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Paul C. Rogers MBChB,

    1. Division of Oncology, Hematology, and Bone Marrow Transplant, British Columbia Children's Hospital and University of British Columbia, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Mary L. McBride MSc

    Corresponding author
    1. Cancer Control Research Program, British Columbia Cancer Agency, Vancouver, British Columbia, Canada
    2. Department of Health Care and Epidemiology, University of British Columbia, Vancouver, British Columbia, Canada
    • Cancer Control Research Program, British Columbia Cancer Research Centre, British Columbia Cancer Agency, 2-107, 675 West 10th Avenue, Vancouver, British Columbia,V5Z 1L3, Canada===

    Search for more papers by this author
    • Fax: (604) 675-8180


Abstract

BACKGROUND:

Increased survival of patients with childhood cancer has resulted in a growing population of survivors within the education system, many of whom may experience educational difficulties. The current study provides a comprehensive assessment of survivors' educational achievements.

METHODS:

Seven hundred eighty-two childhood cancer survivors from the British Columbia (BC) Cancer Registry who attended BC schools from 1995 to 2004, were compared with a randomly selected comparison group of 8386 BC school children. Grade repetition, standard Foundation Skills Assessments (FSA), graduation-year examinations, and special education designations were compared, and factors that affected survivors' educational outcomes were identified.

RESULTS:

Survivors of central nervous system tumors had statistically significant FSA deficits in numeracy and reading (adjusted odds ratios from 0.2 to 0.5 in various grades); leukemia survivors also had lower FSA scores, although most differences were not statistically significant. Other survivors demonstrated no significant differences in FSA scores. Survivors were significantly more likely than controls to receive special education (32.5% vs 14.1%). Females and those who had received radiation treatment (particularly cranial radiation) were at increased risk for poor educational outcomes.

CONCLUSIONS:

The current results have implications for the management of survivors in the education system to maximize their educational experience. Cancer 2009. © 2009 American Cancer Society.

Advances in medical treatment have increased survival rates dramatically for children who are diagnosed with cancer1 to approximately 80%,2 resulting in a growing population of children in the education system who have survived childhood cancer.3-5 Educational difficulties among survivors have been reported by survivors, parents, and teachers.6-8 Survivors attain lower levels of education,9-12 have increased use of special education services,13-16 and have a greater likelihood of grade repetition.13, 17 There is considerable evidence of adverse neurocognitive late effects,18-21 including lower intelligence quotient scores,22, 23 that may inhibit educational performance.

Younger age at diagnosis and/or treatment7, 14, 24-26; cranial or craniospinal radiotherapy (CRT) and/or intrathecal methotrexate (IT MTX)15, 24, 25, 27-29; higher therapeutic doses of CRT or IT MTX17, 24; and diagnoses of central nervous system (CNS) tumors, leukemia (particularly acute lymphoblastic leukemia [ALL]), Hodgkin lymphoma, or neuroblastoma13, 15, 30 as well as being female,30-33 of lower socioeconomic status (SES),20 and having a longer time since diagnosis24 have been reported to be associated with an increased risk of poorer educational outcomes. Previous studies have not always assessed the potential confounding effects of missed school,34 SES,35 and physical health or physical impairments.36 Some studies have included inappropriate comparison groups, or no comparison group,23, 37 or small sample sizes.38 Studies of educational effects commonly have used self-reported or proxy-reported data,13, 30 which are subject to reporting bias.34 To our knowledge, only 2 studies,4, 5 both population-based, have investigated educational achievement using standardized achievement tests. Both survivors of CNS tumors and leukemia survivors had lower ninth grade results than controls, and radiation treatment was a major factor.4, 5

The Childhood/Adolescent/Young Adult Cancer Survivors (CAYACS) Research Program has assembled a cohort of all 5-year survivors of a cancer or tumor diagnosed before age 25 years in British Columbia (BC) from 1970 and population sample comparison groups from provincial registries and has linked these records to provincial administrative datasets. Within this program, our objectives for this study were to compare survivors' educational outcomes, including multiple standardized tests, with outcomes in the general student population, and to identify factors associated with poor achievement.

MATERIALS AND METHODS

For this analysis, 1096 survivors who were diagnosed at age <15 years were selected from the BC Cancer Registry if they had a first primary diagnosis of cancer between 1975 and 1995 and were included in the International Classification of Childhood Cancer39, if they were residents of BC at the time of diagnosis, and if they had survived for ≥5 years since diagnosis. Survivor records were linked by Edudata Canada (a repository of the BC Ministry of Education [MOE] datasets for research use) to BC education records from Kindergarten through Grade 12 (secondary school graduation year) between the 1995/1996 and 2004/2005 school years using personal information (ie, name, date of birth, and sex); 71% of eligible patients were linked successfully, resulting in 782 survivors for analysis. A comparison group of 8386 randomly selected students (approximately 10 times the size of the eligible survivor group) was provided from MOE records with the same distribution of sex and birth year as the eligible survivors.

Data included annual grade level, standardized scores, and percentages of Foundation Skills Assessments (FSAs) written in Grades 4, 7, and 10 in the subject areas of reading, writing, and numeracy; Grade 12 provincial examination scores in English, mathematics (math), biology, history, and communications; and special education program designation. FSAs are province-wide, standardized tests that assess student performance with respect to basic academic skills. FSA participation was calculated as having FSA test data (ie, possessing a scaled score) at the appropriate grade for the corresponding FSA (ie, in Grade 4 and wrote a Grade 4 FSA). Achievement on FSAs was investigated by comparing the proportion of students who did versus students who did not meet or exceed expectations. FSA data were available for Grades 4 and 7 for the 1999/2000 to 2004/2005 school years and for Grade 10 for the 1999/2000 to 2002/2003 school years. Grade 12 provincial examinations are standardized tests for which percentage scores are given. A letter grade mark of C (equal to 60%) was used as an achievement cutoff point for Grade 12 provincial examinations; this mark is considered representative of satisfactory performance with respect to standard learning objectives.40 For Grade 12 courses, participation was determined as those who ever had a school mark (assigned for the course by the school) or provincial examination mark and had reached Grade 12 or graduated. Special education designations refer to special education services/programs that the student required for a disability of an intellectual, physical, sensory, emotional, or behavioral nature; a learning disability; or special gifts or talents.41 Only 1 special education code is recorded per year. Survivors were followed from the later of September 1995 or the beginning of the school year in which the 5-year anniversary of their diagnosis was attained until the earlier of graduation, June 2004, or the last record of registration in the Edudata files. Individuals in the comparison group were followed from the later of September 1995 or their first year of registry in Edudata until the earlier of graduation, June 2004, or their last record of registration. Outcomes that occurred at least 5 years after diagnosis, including the school year in which the 5-year anniversary was attained, were included.

Sociodemographic status and urban/rural status of residence at the start of follow-up were derived using methodology from Statistics Canada,42 which links individuals' postal code of residence to neighborhood-specific income quintiles based on the average income-per-person equivalent. Information on diagnosis was obtained from the BC Cancer Registry, and treatment information was obtained through medical chart abstraction.

Statistical Analysis

Sociodemographic characteristics of the survivors and the comparison group were compared using chi-square tests. For each educational achievement measure, a binary indicator was created (ie, met expectations vs did not meet expectations). Logistic regression was used to compare FSA and Grade 12 course participation rates and educational outcomes of survivors and comparators while adjusting for the effect of potential modifiers. Odds ratios (ORs), adjusted ORs (ORadj), 95% confidence intervals (95% CIs), and P values were determined. All statistical analyses were calculated using the software package R.43

RESULTS

Study Group Characteristics

In total, 782 survivors of childhood cancer and a sample of 8386 individuals from the student population were available for analysis. Both groups were observed for an average of 5.5 school years, and similar proportions (59% of survivors; 57.6% of the student sample) had reached Grade 12.

The distribution of most sociodemographic characteristics was similar between survivors and the student population; however, fewer survivors lived in either rural areas or large metropolitan areas (Table 1). Leukemia was the most common diagnosis among survivors followed by CNS tumors (Table 2). Survivors were a mean age of 4.6 years at diagnosis, and chemotherapy was the most common treatment received (68%).

Table 1. Sociodemographic Characteristics of Survivors and Controls
CharacteristicSurvivors (n=782)Controls (n=8386)P
No.%No.%
  • SES indicates socioeconomic status.

  • *

    SES quintile was missing for 66 survivors and 742 controls, and urban/rural status was missing for 63 survivors and 704 controls.

Sex    .577
 Men43155.1452954 
 Women35144.9385746 
Birth era    .970
 1972-197914017.9151618.1 
 1980-198426533.9290134.6 
 1985-198924030.7251630 
 1990-199513717.5145317.3 
Reached Grade 12    .469
 Yes46159483057.6 
SES quintile*    .101
 1 (Lowest)10214.2135817.8 
 215421.5143318.7 
 314219.8156920.5 
 415621.8160821 
 5 (Highest)16222.6167621.9 
Urban/rural status*    .003
 Rural12617.5142618.6 
 Small community11616.1105013.7 
 Large community12917.9108014.1 
 Metropolitan34848.4412653.7 
Table 2. Clinical Characteristics of Survivors
Treatment*Total No.ChemoNo. of Survivors (%)
ChemoRTCombination
IT ChemoIT MTXRTCRTChemo + RT
  • Chemo indicates chemotherapy; RT, radiotherapy; IT, intrathecal; MTX, methotrexate; CRT, cranial RT; CNS, central nervous system.

  • *

    Categories were not mutually exclusive; data were missing for 27 survivors.

All survivors782536 (68.5)292 (37.3)273 (34.9)227 (29)149 (19.1)181 (23.1)
 Leukemia270270 (100)254 (94.1)240 (88.9)85 (31.5)80 (29.6)85 (31.5)
 Lymphoma5857 (98.3)32 (55.2)27 (46.6)7 (12.1)2 (3.4)7 (12.1)
 CNS16633 (19.9)0 (0)0 (0)58 (34.9)57 (34.3)22 (13.3)
 Neuroblastoma4819 (39.6)0 (0)0 (0)10 (20.8)0 (0)7 (14.6)
 Others240157 (65.4)6 (2.5)6 (2.5)67 (27.9)10 (4.2)60 (25)
Diagnosis at ages birth to 1 y224137 (61.2)54 (24.1)49 (21.9)48 (21.4)20 (8.9)36 (16.1)
 Leukemia5353 (100)51 (96.2)46 (86.8)16 (30.2)15 (28.3)16 (30.2)
 Lymphoma22 (100)1 (50)1 (50)0 (0)0 (0)0 (0)
 CNS306 (20)0 (0)0 (0)6 (20)5 (16.7)1 (3.3)
 Neuroblastoma3511 (31.4)0 (0)0 (0)6 (17.1)0 (0)3 (8.6)
 Others10465 (62.5)2 (1.9)2 (1.9)20 (19.2)0 (0)16 (15.4)
Diagnosis at ages 2-4 y280217 (77.5)143 (51.1)136 (48.6)95 (33.9)67 (23.9)77 (27.5)
 Leukemia136136 (100)132 (97.1)126 (92.6)37 (27.2)36 (26.5)37 (27.2)
 Lymphoma1414 (100)8 (57.1)7 (50)3 (21.4)1 (7.1)3 (21.4)
 CNS5414 (25.9)0 (0)0 (0)27 (50)27 (50)10 (18.5)
 Neuroblastoma118 (72.7)0 (0)0 (0)4 (36.4)0 (0)4 (36.4)
 Others6545 (69.2)3 (4.6)3 (4.6)24 (36.9)3 (4.6)23 (35.4)
Diagnosis at ages 5-14 y278182 (65.5)95 (34.2)88 (31.7)84 (30.2)62 (22.3)98 (35.3)
 Leukemia8181 (100)71 (87.7)68 (84)32 (39.5)29 (35.8)32 (39.5)
 Lymphoma4241 (97.6)23 (54.8)19 (45.2)4 (9.5)1 (2.4)4 (9.5)
 CNS8213 (15.9)0 (0)0 (0)25 (30.5)25 (30.5)11 (13.4)
 Neuroblastoma20 (0)0 (0)0 (0)0 (0)0 (0)0 (0)
 Others7147 (66.2)1 (1.4)1 (1.4)23 (32.4)7 (9.9)21 (29.6)

Educational Outcomes

The study groups had similar levels of grade repetition (21.5% among survivors; 22% among controls) and FSA participation rates (at least 95% of enrollees for all 9 examinations). Survivors were less likely to enroll in English 12 (66.6% vs 79.5%; ORadj, 0.73 [95% CI, 0.58-0.91]) and Math 12 (25.4% vs 37.6%; ORadj, 0.58 [95% CI, 0.46-0.73]) and were more likely to enroll in Communications 12, an alternative to English 12 for those not pursuing postsecondary education (17.8% vs 13.1%; ORadj, 1.30 [95% CI, 0.99-1.71]), than the student sample.

Results of comparisons of education achievement and rates of special education designations are presented in Table 3. Survivors were significantly less likely than comparators to meet or exceed expectations on numeracy in Grades 4 and 7 (ORadj, 0.63 [95% CI, 0.43-0.94] and ORadj, 0.69 [95% CI, 0.51-0.95], respectively), and reading Grade 7 (ORadj, 0.73; 95% CI, 0.53-0.99). There were no significant differences in achievement on any Grade 12 provincial examination.

Table 3. Educational Outcomes of Survivors Versus Comparison Group
OutcomeOverallPMenPWomenP
Comparison Group*CasesORadj (95% CI )Comparison Group*CasesORadj (95% CI)Comparison Group*CasesORadj (95% CI)
No.%No.%%%%%
  • ORadj indicates adjusted odds ratio; 95% CI, 95% confidence interval; FSA, Foundation Skills Assessment.

  • *

    Reference group.

  • Adjusted for sex, urban/rural status, and socioeconomic status quintile. Achievement outcomes were also adjusted for special education designation.

  • Compared with those who did not meet expectations and those who did not write.

  • §

    Compared with those who had a mark below C (60%) and those who did not write.

FSA score meets expectations
 Numeracy
  Grade 483668.97254.10.63 (0.43-0.94).02269.747.80.54 (0.31-0.93).0256860.90.78 (0.43-1.42).423
  Grade 7125566.710754.30.69 (0.51-0.95).02167.454.60.75 (0.49-1.15).19365.8540.64 (0.40-1.01).057
  Grade 10102454.98949.20.89 (0.65-1.22).46555.456.31.17 (0.75-1.82).47954.341.20.63 (0.40-1.03).067
 Reading
  Grade 479965.87757.90.86 (0.58-1.27).45363.950.70.75 (0.44-1.28).2926862.60.95 (0.53-1.71).865
  Grade 7123365.510754.30.73 (0.53-0.99).04361.648.20.70 (0.46-1.06).08869.862.10.78 (0.48-1.25).302
  Grade 10102955.285470.80 (0.58-1.10).16349.4491.08 (0.70-1.67).72161.944.70.57 (0.36-0.91).017
 Writing
  Grade 4100682.910276.70.94 (0.59-1.51).79679.172.60.95 (0.52-1.74).85987.281.30.91 (0.42-1.99).822
  Grade 7141975.413568.50.92 (0.65-1.29).62269.463.61.01 (0.65-1.58).94782.174.70.77 (0.45-1.33).351
  Grade 1012126511764.61.17 (0.83-1.64).36458.163.51.36 (0.86-2.14).18372.965.90.93 (0.56-1.53).772
Grade 12 course enrollment
 English366575.930766.60.73 (0.58-0.91).00670.564.90.86 (0.63-1.17).33581.968.60.58 (0.41-0.81).002
 Mathematics181837.611725.40.58 (0.46-0.73)<.00138.726.30.61 (0.44-0.84).00236.524.30.54 (0.38-0.78).001
 History967207416.10.84 (0.67-1.10).20819.319.11.13 (0.80-1.60).50220.812.40.55 (0.35-0.87).010
 Biology152131.513829.90.97 (0.77-1.22).7642122.71.10 (0.78-1.55).5794338.60.87 (0.64-1.18).366
 Communications63213.18217.81.30 (0.99-1.71).05817.120.71.22 (0.86-1.72).2588.714.31.43 (0.92-2.24).116
Grade 12 provincial examination C (60%) or above§
 English282577.123977.91.08 (0.79-1.47).64071.670.60.92 (0.62-1.36).66582.386.11.45 (0.85-2.47).169
 Mathematics129671.37765.80.74 (0.48-1.14).17368.762.20.69 (0.39-1.21).19574.370.60.78 (0.38-1.58).489
 History69571.95675.71.13 (0.63-2.01).68072.677.11.24 (0.60-2.55).55771.173.10.96 (0.36-2.55).930
 Biology92961.18561.60.90 (0.62-1.32).59858.559.70.88 (0.48-1.63).68362.5630.92 (0.56-1.51).744
 Communications47274.75769.50.94 (0.54-1.63).82873.273.11.34 (0.65-2.79).42977.963.30.50 (0.20-1.26).141
 Special education117614.125432.53.05 (2.58-3.60)<.00115.934.12.89 (2.32-3.61)<.0011230.53.31 (2.57-4.27)<.001
 Learning disability2272.7303.81.44 (0.97-2.13).06844.61.22 (0.76-1.96).4141.32.92.25 (1.12-4.50).022
 Physical disability941.115019.221.47 (16.33-28.22)<.0011.318.818.96 (13.2-27.23)<.001119.726.54 (17.35-40.59)<.001
 Visual disability70.1232.938.55 (16.43-90.48)<.0010.12.828.99 (10.06-83.54)<.0010.13.169.86 (15.26-319.89)<.001
 Hearing disability240.3435.59.69 (5.35-17.54)<.0010.32.39.11 (3.89-21.38)<.0010.33.110.35 (4.50-23.79)<.001

In total, 254 (33%) of the 782 survivors had been designated to receive special education, including 150 survivors (19%) who were designated for special education because of a physical disability. Survivors were 3 times more likely to have a special education designation than the student sample (ORadj, 3.05; 95% CI, 2.6-3.6). In particular, survivors had more physical, visual, and hearing disability designations (ORadj, 21.47 [95% CI, 16.3-28.2], ORadj, 16.18 [95% CI, 10.1-25.9], and ORadj, 9.69 [95% CI, 5.4-17.5], respectively). There were no significant differences in the rates of learning disability or gifted designations.

Girl survivors were more at risk for poor educational achievement and were more likely to have a learning disability than control girls, whereas survivors who were boys were no more likely than boys in the comparison sample to have a learning disability (Table 3). Within the general student sample, there were significantly fewer girls than boys with a learning disability (ORadj, 0.30 [95% CI, 0.2-0.4]); in contrast, among survivors, there was no significant difference in learning disability designations between boys and girls (ORadj, 0.56 [95% CI, 0.3-1.2]).

Diagnosis-specific educational outcomes are presented in Table 4. Only survivors of CNS tumors experienced significant excess risk for poor educational outcomes, with ORs ranging from 0.20 to 0.62 for FSA assessments and >6 times the rate of special education enrollees (ORadj, 6.1 [95% CI, 4.4-8.5]). Leukemia and neuroblastoma survivors showed no significant differences on any educational achievement measure compared with the student sample but were significantly more likely to have special education (for leukemia: ORadj, 3.1 [95% CI, 2.3-4.0]; for neuroblastoma: ORadj, 2.3 [95% CI, 1.2-4.3]). Students who received special education had worse educational outcomes whether or not they were survivors (Table 5), and the achievement differential between those with and without a special education designation was similar for both groups.

Table 4. Educational Outcomes by Diagnosis
OutcomeControls, n=8386*Leukemias, n=270CNS Tumors, n=166Neuroblastomas, n=48Others, n=298
%%ORadj (95% CI)P%ORadj (95% CI)P%ORadj (95% CI)P%ORadj (95% CI)P
  • CNS indicates central nervous system; FSA, Foundation Skills Assessment; ORadj, adjusted odds ratio; 95% CI, 95% confidence interval; NA, not applicable.

  • *

    Reference group.

  • Adjusted for sex, urban/rural status, and socioeconomic status quintile. Achievement outcomes also were adjusted for special education designation.

  • †, ‡

    Compared with those who did not meet expectations and those who did not write.

  • §

    Compared with those who had a mark below C (60%) and those who did not write.

FSA score meets expectations
 Numeracy
  Grade 468.953.10.56 (0.30-1.04).06831.30.20 (0.06-0.62).00570.61.23 (0.41-3.70).70756.90.79 (0.43-1.45).446
  Grade 766.755.40.67 (0.41-1.10).11038.20.40 (0.19-0.85).01768.41.05 (0.38-2.92).92257.10.81 (0.49-1.33).400
  Grade 1054.954.90.77 (0.46-1.28).30826.50.37 (0.17-0.82).015702.13 (0.53-8.60).28859.71.28 (0.78-2.09).330
 Reading
  Grade 465.855.10.74 (0.40-1.38).34629.40.22 (0.07-0.68).00864.71.02 (0.36-2.85).97766.71.40 (0.75-2.63).291
  Grade 765.554.10.68 (0.42-1.10).11538.20.43 (0.21-0.88).02257.90.73 (0.28-1.87).50861.40.95 (0.57-1.58).856
  Grade 1055.249.20.84 (0.51-1.40).50732.40.47 (0.22-0.98).045702.29 (0.58-9.12).23848.60.82 (0.51-1.33).429
 Writing
  Grade 482.981.61.34 (0.58-3.06).49156.30.26 (0.09-0.77).01576.50.72 (0.22-2.33).58878.41.18 (0.57-2.47).656
  Grade 775.470.30.94 (0.54-1.62).81252.90.58 (0.27-1.21).14573.40.94 (0.32-2.78).91172.91.16 (0.65-2.06).453
  Grade 106572.31.59 (0.90-2.81).11347.10.62 (0.30-1.26).186601.11 (0.29-4.18).87866.71.22 (0.72-2.06).453
Grade 12 course enrollment
 English75.969.70.80 (0.55-1.16).24151.40.41 (0.26-0.63)<.00176.51.32 (0.40-4.31).64971.70.90 (0.63-1.30).585
 Mathematics37.625.20.56 (0.38-0.83).00417.10.33 (0.18-0.59)<.00147.11.45 (0.53-4.00).47328.30.70 (0.49-0.99).047
 History2021.91.20 (0.80-1.80).3768.60.46 (0.23-0.92).02817.71.01 (0.28-3.58).98815.20.76 (0.49-1.17).212
 Biology31.533.61.10 (0.76-1.59).61222.90.54 (0.31-0.91).02247.13.23 (1.18-8.85).02329.41.01 (0.71-1.43).977
 Communications13.115.51.17 (0.74-1.86).49122.91.63 (0.98-2.72).06017.71.05 (0.29-3.80).94316.91.29 (0.84-1.98).241
Grade 12 provincial examination grade C (60%) or above§
 English77.176.90.96 (0.58-1.56).85975.90.92 (0.45-1.89).82961.50.41 (0.13-1.35).14381.11.46 (0.88-2.42).147
 Mathematics71.371.80.85 (0.40-1.79).67161.10.84 (0.24-2.78).77362.50.54 (0.12-2.46).42363.50.68 (0.36-1.28).229
 History71.976.51.12 (0.49-2.56).78655.60.45 (0.12-1.70).23666.70.87 (0.80-9.91).90782.11.92 (0.65-5.72).240
 Biology61.161.50.91 (0.50-1.67).76162.50.63 (0.24-1.61).33162.51.20 (0.28-5.15).80361.10.96 (0.53-1.72).879
 Communications74.783.32.01 (0.66-6.11).218500.44 (0.17-1.11).081100NA 710.89 (0.38-2.09).794
 Special education14.1333.06 (2.34-3.99)<.001476.11 (4.40-8.49)<.00129.22.29 (1.21-4.32).01124.52.06 (1.56-2.72)<.001
 Learning disability2.75.62.11 (1.22-3.63).0071.80.69 (0.22-2.21).5376.31.94 (0.59-6.35).27531.14 (0.57-2.25).713
 Physical disability1.114.815.75 (10.57-23.47)<.00137.356.94 (38.62-83.94)<.00110.49.46 (3.62-24.71)<.00114.415.34 (10.42-22.59)<.001
 Visual disability0.41.519.58 (5.63-68.12)<.0016.695.46 (35.82-254.35)<.0010  2.734.53 (12.32-96.76)<.001
 Hearing disability0.30 4.216.0 (6.71-37.92)<.0010  4.717.66 (9.00-34.68)<.001
Table 5. Effect of Special Education on Educational Achievement Among Survivors and Controls
OutcomeCasesOR (95% CI)PControlsOR (95% CI)PInteraction P
Special EducationNo Special Education*Special EducationNo Special Education*
Total No.% With OutcomeTotal No.% With OutcomeTotal No.% With OutcomeTotal No.% With Outcome
  • OR indicates odds ratio; 95% CI, 95% confidence interval; FSA, Functional Skills Assessment.

  • *

    Reference group.

  • Compared with those who did not meet expectations and those who did not write.

  • Compared with those who had a mark below C (60%) and those who did not write.

FSA score meets expectations
 Numeracy
  Grade 450328367.50.20 (0.08-0.53).00120746.4100773.50.30 (0.22-0.41)<.001.564
  Grade 78740.211065.50.28 (0.14-0.55)<.00141651146771.10.41 (0.33-0.52)<.001.734
  Grade 107530.710662.30.27 (0.13-0.56)<.00140543146058.40.55 (0.44-0.69)<.001.038
 Reading
  Grade 450388369.90.21 (0.08-0.56).00220746.4100769.80.38 (0.28-0.52)<.001.413
  Grade 78733.311070.90.16 (0.08-0.32)<.00141651.4146769.50.48 (0.38-0.60)<.001.014
  Grade 107532010657.60.25 (0.12-0.52)<.00140543.5146058.40.59 (0.47-0.74)<.001.106
 Writing
  Grade 450528391.60.04 (0.01-0.18)<.00120763.8100786.80.28 (0.20-0.40)<.001.037
  Grade 78756.311078.20.32 (0.15-0.64).00141654.8146781.20.29 (0.23-0.37)<.001.374
  Grade 10754810676.40.29 (0.14-0.59).00140549.1146069.40.46 (0.37-0.58)<.001.118
Grade 12 provincial examination grade C (60%) or above
 English507625778.20.82 (0.36-1.86).63231280.1335076.81.25 (0.93-1.68).147.249
 Mathematics2272.79564.21.61 (0.39-6.65.50917287.2164469.63.02 (1.90-4.81)<.001.316
 History15805974.68.93 (0.88-90.35).0649481.987270.81.87 (1.08-3.25).025.674
 Biology265011264.30.65 (0.22-1.90).43013970.5138160.11.68 (1.15-2.47).008.059
 Communications3669.44669.60.61 (0.13-2.82).52810669.852675.70.79 (0.50-1.27).333.961

Treatment Factors

Survivors who received radiation therapy, particularly CRT, were significantly more likely to have physical and hearing disabilities and poor educational achievement compared with those who had not received radiation (Table 6). CNS survivors who received CRT had the poorest FSA outcomes, and <20% of this group met expectations on all 9 FSA examinations. Survivors who received chemotherapy were more likely to have a physical disability compared with those who received no chemotherapy (ORadj, 2.81 [95% CI, 1.46-5.42]; P = .002). Survivors who received IT MTX were more likely to have special education designations than those who did not receive IT MTX (Table 5). No significant differences were observed among survivors who were diagnosed at different ages (Table 6).

Table 6. Treatment Factors Related to Selected Educational Outcomes Among Survivors
OutcomeIT MTX vs No IT MTX*RT vs No RT*CRT vs No CRT*
ORadj (95% CI)PORadj (95% CI)PORadj (95% CI)P
  • IT indicates intrathecal; MTX, methotrexate; RT, radiotherapy; CRT, cranial RT; ORadj, adjusted odds ratio; 95% CI, 95% confidence interval; FSA, Functional Skills Assessment.

  • *

    Reference group.

  • Adjusted for sex, urban/rural status, and socioeconomic status quintile.

  • Compared with those who did not meet expectations and those who did not write.

  • §

    Compared with those who had a mark below C (60%) and those who did not write.

FSA score meets expectations
 Numeracy
  Grade 40.31 (0.05-2.01).2220.18 (0.04-0.78).0180.27 (0.04-1.69).160
  Grade 70.33 (0.08-81.31).1160.51 (0.22-1.16).1070.50 (0.17-1.52).222
  Grade 100.11 (0.02-0.62).0120.47 (0.19-1.13).0910.20 (0.06-0.67).009
 Reading
  Grade 40.75 (0.15-4.89).7640.92 (0.24-3.55).8992.42 (0.43-13.44).314
  Grade 70.45 (0.12-1.76).2540.69 (0.30-1.55).3660.51 (0.16-1.60).249
  Grade 100.22 (0.05-0.97).0460.86 (0.36-1.87).6460.32 (0.11-0.94).039
 Writing
  Grade 42.02 (0.18-22.48).5660.56 (0.10-3.10).5031.18 (0.11-12.79).891
  Grade 70.12 (0.02-0.83).0320.65 (0.29-1.47).3020.27 (0.09-0.83).022
  Grade 100.29 (0.05-1.54).1450.81 (0.36-1.82).0650.43 (0.16-1.15).092
Grade 12 provincial examination mark of C (60%) or above§
 English0.48 (0.13-1.75).2640.63 (0.31-1.28).2030.63 (0.27-1.47).284
 Mathematics0.84 (0.14-4.95).8510.88 (0.3-2.62).8220.66 (0.15-2.90).578
 History0.71 (0.03-17.94).83811.14 (1.15-107.69).0375.21 (0.49-55.78).172
 Biology0.08 (0.01-0.68).0210.77 (0.31-1.95).5830.58 (0.19-1.82).352
 Communications1.27 (0.06-25.22).8770.24 (0.05-1.15).0740.25 (0.05-1.30).099
 Special education0.66 (0.34-1.31).2361.03 (0.72-1.48).8681.09 (0.71-1.69).681
 Learning disability6.57 (0.94-44.33).0530.72 (0.29-1.79).4830.91 (0.31-2.65).859
 Physical disability1.42 (0.54-3.75).4751.90 (1.24-2.90).0032.00 (1.23-3.24).005
 Hearing disability2.53 (0.21-31.11).4694.10 (1.38-12.14).0116.94 (1.88-25.6).004
 Visual disability0.11 (0.01-0.80).0291.35 (0.51-3.55).5431.20 (0.41-3.56).741

DISCUSSION

In this study, we examined a large set of educational outcomes, including standardized achievement tests, in a geographically defined survivor cohort that included all childhood cancer diagnoses. In our study, increased risks for poor educational achievement were limited to survivors of CNS tumors; the results for leukemia survivors were not statistically significant. Other studies that have assessed different populations with fewer outcome measures have reported an excess risk of poor educational achievement among survivors of CNS tumors and leukemia.9, 13, 15

The reciprocal pattern of enrollment of survivors in English and Communications 12 and the reduced enrollment of survivors in Math 12, coupled with the lack of difference in Grade 12 examination results between survivors and the student sample, suggest that there may be self-selection for ability in these subjects. In particular, the findings for Math 12 most likely reflect difficulties in mathematic skills; mathematic deficits and poor arithmetic achievement have been reported previously among childhood cancer survivors.20, 44, 45 In contrast, the high participation rates in FSA tests in the current study suggest that the results of these tests provide comprehensive, valid assessments of achievement for comparison with the general student population.

Consistent with earlier research,14-16, 30 survivors were more likely than the student sample to be enrolled in special education programs as a consequence of the types of disabilities expected among survivors, in particular for survivors of CNS tumors and leukemia.33, 37, 46-49 Unfortunately, the administrative dataset we used did not contain information on specific disabilities, so we were unable to examine relations between specific problems and educational outcome in this analysis.

Among the diagnostic groups, CNS tumor survivors had the lowest enrollment rates in all observed Grade 12 subjects except communications. The consistently poor FSA scores across all subjects and grades suggest that this group is dealing with multiple learning deficits that are not ameliorated over time. Leukemia survivors had poorer results for numeracy and reading tests than for writing. Unlike the results reported by Barrera et al,13 neuroblastoma survivors in our study were not more likely to have academic problems; however, in agreement with Mitby et al,15 they did use special education services more often. Unfortunately, in our study there, were insufficient numbers of survivors in diagnostic subgroups or with multiple test results to explore other diagnostic subgroup results or trends with longer time since diagnosis.

CRT is associated with poor educational outcomes among survivors of childhood cancer,24, 27, 28 as we also demonstrated in the current study. Reports have indicated that IT MTX is a risk factor for poor cognitive performance and neurologic function.34, 50 However, in our study, survivors who received IT MTX did not have lower educational achievement. Survivors who received IT MTX were more likely to have a learning disability, and the majority of these individuals were survivors of leukemia.

Younger age at diagnosis has been considered a risk factor for cognitive deficits.7, 51, 52 However, we observed no statistically significant differences in educational achievement among survivors aged <2 years at diagnosis compared with survivors who were diagnosed at age ≥5 years.

The sex differences we observed among survivors were similar to those reported in other studies.30, 31, 33 Being a girl also appears to be a risk factor for educational difficulties in those with disorders such as autism and attention deficit hyperactivity disorder; although these disorders are more prevalent among boys, girls are impaired more severely,53-55 and in particular have more impaired intellectual functioning.56-58

To our knowledge, this is the first population-based cohort study to use standardized measures to examine educational late effects of survivors of all childhood cancers. By using a geographically defined cohort, a randomly sampled comparison group, and standardized achievement measures from administrative data, we were able to minimize potential bias due to incomplete or nonrepresentative sampling and recruitment, loss to follow-up, and self-report and recall. Reasons for nonlinkage in our study included age older than the standard school age range (only 30% of survivors who were diagnosed at ages 12-14 years were linked) and inability to match name/birthdate information.

Limitations of this study include the lack of statistical power to examine some subgroups, risk factors, and outcomes; the lack of information on educational achievement in grades for which no standardized test is given; the lack of information on absenteeism as a factor; and the existence of only 1 special education code per individual per year, which leads to likely under-reporting of some types of special education designations. The results from this study and others have important implications for survivors and their parents, clinicians, and educators, all of whom need to be aware of at-risk groups, potential educational difficulties, and associated risk factors in order to meet long-term educational needs.19 Sharing of risk information between clinicians, parents, and school personnel is fundamental59 in addressing the transition to school. Early identification of problems16, 50 and regular monitoring of progress over time36 in the school system are essential to provide appropriate special education services or approaches to learning. The relation between special education programs and achievement among survivors (other than survivors of CNS tumors or leukemia) needs to be explored further to address the question of the contribution of special education to achievement in this group. Finally, although studies consistently have indicated that survivors experience adverse neurocognitive late effects,18-21, 50 these need to be linked clearly to poor achievement in school60 and educational intervention opportunities.61

Acknowledgements

We gratefully acknowledge cooperation in allowing access, use, and linkage of the data to support this program from the following: British Columbia Cancer Registry, British Columbia Cancer Agency, British Columbia Children's Hospital, the British Columbia Ministries of Health and Education, the Centre for Health Services and Policy Research, and Edudata Canada at the University of British Columbia.

Conflict of Interest Disclosures

Supported by the CAYACS Program, funded by Canadian Cancer Society Program Project Grant 16001.

Ancillary