Socio-economic Status Plays Important Roles in Childhood Cancer Treatment Outcome in Indonesia

The cure rate of childhood cancer may be as high as 80% in high-income countries, but frequently is less than 35% in low-income countries (Smith et al., 2010). Poor treatment adherence by health-care providers, parents and patients importantly contributes to this difference in survival. Abandonment of treatment, a severe form of non-adherence, is one of the most common reasons for treatment failure in low-income countries (Arora et al., 2010; Lam et al., 2012) Recently the SIOP PODC Abandonment of Treatment Working Group declared that treatment abandonment can no longer be ignored by the international pediatric oncology community (Mostert et al., 2011). More insight into underlying causes is needed. Providing access to care, improving treatment adherence and reducing abandonment should be a top priority for those involved in childhood cancer care in low-income countries in order


Introduction
The cure rate of childhood cancer may be as high as 80% in high-income countries, but frequently is less than 35% in low-income countries (Smith et al., 2010).Poor treatment adherence by health-care providers, parents and patients importantly contributes to this difference in survival.Abandonment of treatment, a severe form of non-adherence, is one of the most common reasons for treatment failure in low-income countries (Arora et al., 2010;Lam et al., 2012) Recently the SIOP PODC Abandonment of Treatment Working Group declared that treatment abandonment can no longer be ignored by the international pediatric oncology community (Mostert et al., 2011).More insight into underlying causes is needed.Providing access to care, improving treatment adherence and reducing abandonment should be a top priority for those involved in childhood cancer care in low-income countries in order RESEARCH ARTICLE
Although it is generally believed that financial difficulties of parents are a major contributor to abandonment of childhood cancer treatment (Bonilla et al., 2009;Sitaresmi et al., 2010), very few studies have actually demonstrated any impact of parental socio-economic status on abandonment and survival of childhood cancer patients in low-income countries.Also in Prof Dr RD Kandou Hospital (KH) in Manado, Indonesia, insight into the role that socio-economic status plays in childhood cancer treatment is missing.
Here we examined the influence of parental socioeconomic status on childhood cancer treatment outcome, and parental experiences during cancer treatment of their children at KH.The obtained insights may help to design strategies to improve adherence, reduce abandonment and increase survival rates in low-income countries.

Setting
Indonesia has approximately 245 million inhabitants, 27% (66 million) of whom are children younger than 15 years (CIA, 2012).A childhood cancer incidence of 102 per million children less than 15 years old in low-income countries (Parkin et al., 1988;1998;Howard et al., 2008), leads to an estimated 6,700 new childhood cancer cases in Indonesia each year.
Our study was conducted at KH in Manado at the island of Sulawesi.This academic hospital is the only centre treating children with cancer in the provinces of Gorontalo, North-Sulawesi, North-Maluku and West-Papua.The estimated population KH hereby serves is 5 million people, including 1.4 million children less than 15 years old.Around 140 childhood cancer patients under 15 are expected in KH's service area.
The pediatric oncology department at KH covers 18 beds and is run by 2 pediatric oncologists and 17 nurses.Three different wards are recognized: first, second and third class.With rising rank, conditions on ward and quality of provided food improve and number of children per hospital room decreases.However, treatment protocols and medical staff are the same for all patients.
No information material about cancer and its treatment is available.
Indonesia distinguishes three types of social healthinsurance: 1) Askes for civil-servants, 2) Jamsostek for private-sector employees, 3) Jamkesmas for the poor.Jamkesmas was instituted in 2004.Jamkesmas-insured patients do not pay premiums (Rokx et al., 2009).Because health-insurance plans changed several times, it has been unclear which costs were covered.
Children are assigned to a specific hospital ward on basis of parental income and health-insurance.Askes patients are assigned to first or second class, depending on the importance of their parents' position in government.Jamsostek patients are assigned to second class.Jamkesmas patients are automatically assigned to third class.

Study design
This was a combined retrospective medical records study and cross-sectional study with semi-structured questionnaires.
In the medical records study inclusion criteria for all children were newly diagnosed malignancy and age between 0 to 16 years at diagnosis.First presentation of patients occurred between January 1999 and October 2009.
We collected the following variables from medical records: name and registration number of childhood cancer patient, type of cancer, gender, age at first presentation, year of first presentation, assigned hospital class at diagnosis, health-insurance and treatment outcome.
Malignancies were diagnosed by histological examination and leukemias by cytomorphology.Malignancies were classified in 5 groups: 1) hematological tumors, 2) brain tumors, 3) solid tumors, 4) rare tumors, 5) a group "unspecified" which included patients of whom only the originate tissue of malignancy was known, while pathology was lacking.
Treatment outcome was defined as abandonment of treatment, death, progressive or relapsed disease and event-free survival.In line with SIOP recommendations (Mostert et al., 2011), abandonment of treatment was defined as failure to start or continue scheduled curative treatment during 4 or more consecutive weeks.
Parental socio-economic status divided children as coming from poor or prosperous families, based on 2 determinants: 1) Assigned hospital class at diagnosis.Patients attending first and second class were classified as prosperous.Patients attending third class as poor.2) Health-insurance.Askes and Jamsostek were classified as prosperous, Jamkesmas as poor.Both determinants were obtained routinely during hospital admission and recorded in medical records.In case of discordance between both determinants, assigned hospital class at diagnosis prevailed.
The cross-sectional study consisted of a semistructured questionnaire.The focus was to assess psychological and socio-economic experiences of parents during cancer therapy.Participants were caretakers of childhood cancer patients who were hospitalized or visited the outpatient clinic at KH between October 2011 and January 2012.Respondents were interviewed by an independent interviewer.The questionnaire contained statements that parents could evaluate on three-point rating scales (agree/uncertain/disagree).A panel of Dutch and Indonesian doctors and psychologists made sure of appropriate, clear and coherent statements.Questionnaire was pilot-tested on a separate group of parents for its content, clarity of language and cultural sensitivities.A few minor adjustments were made on basis of pilot-test.Anonymity and confidentiality were guaranteed.Study was approved by Medical Ethics Committee of KH.

Data analysis
Frequency distributions were calculated.Relationship between treatment results, parental socio-economic status and patient characteristics were evaluated by chi-squared and Fisher's Exact tests.Probability of event-free survival was estimated by Kaplan-Meier method: estimates were compared using log-rank test.Event-free survival was measured from date when patient received diagnosis of cancer to first treatment failure or last follow-up.Data management and analysis were performed with SPSS for Windows version 17.0.Fisher's Exact tests for variables with more than two categories were performed in R version 2.9.1.

Medical record study
In the period January 1999 to October 2009, 145 patients were diagnosed with a malignancy and met study's inclusion criteria.These 145 patients consisted of 76 (52%) boys and 69 (48%) girls.New patients seen in KH for pediatric cancer ranged from 4-28 per annum.Annual average was 15 patients.Ages at admission ranged between 0-15 years.Incidence peak was observed between DOI:http://dx.doi.org/10.7314/APJCP.2012.13.12.6491 Socio-economic Status Plays Important Role in Childhood Cancer Treatment Outcome in Indonesia 3-6 years.Mean age at time of diagnosis was 6.1 years (SD=3.7),median 5.4.
Of 145 children, treatment outcome was not known in 2 patients.These patients had moved and received treatment in another hospital.
Abandonment of treatment was the most common cause of treatment failure (Figure 2).Of 143 patients, 68 (48%) abandoned treatment: 40 (59%) patients refused therapy and 28 (41%) patients initially started treatment and later dropped-out.Table 1 shows reasons of parents for abandoning therapy of their child.It is noteworthy that this information was obtained from medical records, and not from families concerned.
The least common treatment failure was progressive or relapsed disease, which occurred in 13 (9%) patients: progressive disease (1) and relapse (12).Distribution of progressive or relapsed disease was: during treatment (10), and after completion of treatment (3).
Of 143 patients with documented treatment outcome, 112 patients were classified as poor (78%) and 31 patients as prosperous (22%).No discordance between both determinants of socio-economic status occurred.No significant differences in types of cancer were found between both groups.The most likely treatment outcome in poor patients (n=112) was abandonment of treatment (n=60, 54%), whereas in prosperous patients (n=31) it was both event-free survival and treatment-related death (n=9, 29%).Figure 3 shows significant differences in treatment outcome between poor and prosperous patients (P<0.0001).Odds-ratio for treatment abandonment was 3.3 (95% CI: 1.4-8.1,p=0.006) for poor versus prosperous population.
Of 143 patients with documented treatment outcome, time until event could not be retrieved in 5 charts.Figure 4 shows that event-free survival estimate was significantly higher in prosperous (n=29) than in poor (n=109) patients (P=0.001).Note that Figure 3 shows actual percentages, whereas Kaplan-Meier estimates in Figure 4 show timedependent probability estimates.
Parental education and communication: Most parents (n=38, 95%) would like to receive more information about cancer and its treatment.More contact with their doctor was desired by 39 parents (98%).When the doctor explained about disease, treatment and medication, 16 parents (40%) found it hard to understand.Eighteen parents (45%) had difficulties with learning or memorizing doctor's statements and advices.Fifteen parents (38%) had difficulties understanding vocabulary of doctors.Parents of other children were used as source of information about cancer by 35 caretakers (88%).
Medication adherence: Table 2 lists reasons for not administering drugs at home.It was easier to remember to give medication to their child after they recently visited hospital according to 36 parents (90%).Difficulties to tell their doctor the truth about how they managed their disease and took prescribed medication were mentioned by 10 parents (25%).
Marital status: Of all parents, 39 (98%) were married and 1 parent (3%) was widowed.Difficulties to understand the way their spouse was responding to all stress were experienced by 17 parents (43%).Fifteen parents (38%) had marital problems due to disease of their child.
Siblings: Of 40 families, 32 (80%) had more than one child.Twenty-three parents (72%) had less time and energy for siblings.Feelings of guilt towards siblings were expressed by 14 parents (44%).Siblings felt neglected according to 13 parents (41%).Seventeen parents (53%) were afraid of the well-being of siblings.
Coping: Impact of cancer on family life increased during course of treatment according to 34 families (85%).Parents were preoccupied with fatal return of cancer (n=25, 63%).Parents described their child's emotional condition as unhappy in 5 cases (13%), and themselves as unhappy in 10 cases (25%).Caretakers were positive about quality of relationship with their child (n=34, 85%).Table 3 shows coping characteristics.

Socio-economic aspects
Parental employment: Parents of 19 children (48%) had regular income per month.Income decreased since start of treatment according to 22 families (55%).During  treatment 19 fathers (48%) lost their job, 16 of whom (84%) identified their child's cancer as the cause.In total 4 mothers (10%) lost their jobs, and all (100%) due to disease of their child.
Financial difficulties: Loss of income led to financial difficulties (n=25, 63%) and debts (n=22, 55%).Financial difficulties were perceived as great burden according to 22 parents (55%).Costs of treatment forced 4 parents (10%) to withhold their child of certain aspects of treatment.Only 5 parents (13%) believed that they would not be able to complete treatment due to financial problems, and 6 (14%) parents were uncertain.
Transportation: Table 4 illustrates transportation characteristics.Distances to clinic, absence of proper transport and transportation costs were never reasons to miss hospital appointments according to consecutively 20 (50%), 21 (53%) and 21 (53%) families.

Discussion
Abandonment of treatment was the most important reason for treatment failure at KH in Manado, Indonesia.Because most abandonment occurs during diagnostic process, efforts to reduce its magnitude should be implemented when patients first enter hospital.Differences in treatment outcome between poor and prosperous patients were sizeable.Abandonment of treatment was significantly more frequent in poor patients.Kaplan-Meier estimate of event-free survival was significantly lower in poor patients.
According to medical records the most important reason for abandonment of treatment was financial difficulties of parents.During interviews with caretakers we indeed found that socio-economic impact of cancer treatment was profound, despite availability of healthinsurance.Many parents became unemployed due to illness of their child.Prolonged treatment costs combined with decreased incomes caused most families to experience financial difficulties that resulted in debts.Caretakers indicated that health-insurance covered only a part of medical expenses.
According to medical records another important reason for abandonment of treatment was disbelief in usefulness of chemotherapy.From questionnaires we learned that parents often believed that health of their child was beyond doctor's control and determined by luck, fate or God.Therefore conventional treatment may not be considered sufficient to cure their children (McLean et al., 2006;Hamidah et al., 2009;Al-Qudimat et al., 2011).Spiritual or religious help and traditional medicine were often consulted.
Parents may not believe in effectiveness of conventional cancer treatment because parental education by healthcare providers is lacking (Yeh et al., 1999).Almost all parents would have liked to receive more information and attention from their doctor.Many parents have difficulties understanding the information, advices and vocabulary of doctors.As a result, parents may misapprehend the need to complete treatment.Adherence of parents increases if usefulness of treatment and risks of non-adherence are understood (Yeh et al., 1999;Mostert et al., 2010).
In Manado, we also found that beliefs, attitude and behavior of health-care providers differ between poor and prosperous families.Doctors expect poor families to be less adherent and unable to complete treatment.Subsequently less elaborate information, time and attention is given to poor parents (Mostert et al., Submitted).This lack of information may contribute to higher rates of abandonment seen in the poor.
Family life changes dramatically when a child is diagnosed with cancer (Stam et al., 2006).Parents had most difficulties coping with their child's disease and some felt depressed.Parents of other children with cancer were frequently consulted for information and advice.Development of a parent-organization could be beneficial (Naafs-Wilstra et al., 2001).
The referral area of KH suggests that the number of patients with cancer is many times larger than the annual average of 15 patients who visit the pediatric oncology department.Causes in low-income countries are numerous.Lacking awareness of health-issues may delay seeking medical attention.Parents may not recognize symptoms or believe cancer is incurable and medical help futile (Howard et al., 2008;Mostert et al., 2012).Parents may rely on alternative treatment or have no transportation to reach hospital.Families may not afford treatment (Sitaresmi et al., 2010).Health-care providers may not be skilled to recognize childhood cancer.Laboratory and imaging devices may be lacking (Howard et al., 2008).Personnel may believe cancer is incurable or that parents cannot pay treatment and therefore decide referral is not worthwhile (Mostert et al., 2012).
Similarities were found with our previous research on childhood cancer treatment in Yogyakarta, Indonesia (Mostert et al., 2006;2008).This is all the more  remarkable because Indonesia, the largest archipelago in the world, only became one unified country after gaining independence from Dutch colonial rule in 1949.Differences between Yogyakarta on Java and Manado on Sulawesi are therefore distinct.Distance between them is about 2000 km and both isles have a different history, ethnic population, native language, culture and religion.Inhabitants of Yogyakarta are predominantly Muslim and in Manado Christian (Ricklefs, 2001;CIA, 2012).
Our study has some limitations.Time periods of medical records study and questionnaire study differed.Parents of children who had abandoned treatment were not interviewed.Although the children of the interviewees are still at risk to abandon treatment, this may mean that more adherent caretakers participated in the questionnaire study.
Outcome of childhood cancer treatment could be improved substantially by interventions that help to prevent abandonment of treatment.On the basis of our retrospective and cross-sectional study we recommend the following: 1) Comprehensive parental education at diagnosis about cancer, treatment and necessity of treatment adherence, 2) Follow-up system for detecting and contacting promptly families who miss hospital appointment, 3) Communication training of health-care providers.Health-care providers in low-income countries need to realize that technical know-how is insufficient to cure their patients, 4) Program enabling parents to maintain their income level.Handicraft-programs (Howard et al., 2004;AMPATH, 2012) offering parents work opportunities to cover treatment costs are good examples.5) Adapt health-insurance so that all necessary medical expenses are covered.These measures most likely will have a positive influence on adherence and treatment outcome of children with cancer in Sulawesi, Indonesia, and many other low-income countries as well.