1. INTRODUCTION

The resultant consequences of cancer, particularly immunosuppression, makes the presence of coexisting conditions—described as any medical condition that co-occurs with an index condition of interest in an individual—relatively common [1–3]. These conditions may arise from the malignancy itself or the treatment interventions used in the management of cancer [4,5]. For example, the immunosuppression associated with cancer predisposes children to conditions such as infections [6]. Children exposed to anthracyclines and cardiac irradiation are at risk of developing cardiotoxic complications such as valvular abnormalities and pericardial diseases during and after the treatment of cancer [7]. The use of high doses of alkylating agents and platinum compounds have also been identified as risk factors for lung fibrosis, pulmonary pneumonitis, thyroid abnormalities, ototoxicity, and impairment of renal function [8–11].

The risk of cancer, on the other hand, may be increased by the presence of other pre-existing conditions—especially those which are similarly associated with immunosuppression—such as Human Immunodeficiency Virus (HIV) infection [12]. HIV infection is a risk factor for some cancers including Kaposi sarcoma and non-Hodgkin’s lymphoma [12] while hepatitis B virus infection has been linked to hepatocellular carcinoma [13]. Conditions that are characterized by chronic inflammation have also been found to stimulate tumorigenesis [14].

The prevalence of conditions coexisting in adults with cancer, especially chronic conditions such as hypertension, diabetes, peptic ulcer, and other cardiovascular conditions has been described in literature [15–18]. There is, however, a paucity of information on the prevalence of coexisting conditions in children with cancer in South Africa. Cognizance of coexisting conditions in patients with cancer is important due to their possible influence on treatment decisions—by the modification of treatment protocols—and the consequent impact on treatment outcome [1,15,19–21]. This study, therefore, aimed at identifying conditions that coexist with cancer in children and adolescents undergoing cancer chemotherapy in the South African private health sector.

2. MATERIALS AND METHODS

3. RESULTS

4. DISCUSSION

This study aimed at identifying coexisting conditions in children and adolescents on treatment for cancer using the ICD-10 codes recorded on claims for non-cytotoxic medicines. In the absence of specific ICD-10 codes, the main pharmacological classes claimed were analyzed. The results of this study indicate that specific ICD-10 codes were available for only 13.65% (N = 2631) of the total medicine claims over the study period. Non-specific ICD-10 codes indicated on the majority of medicine claims included those for repeat prescriptions (Z76.0), encountering health services in unspecified conditions (Z76.9), and failure of patient or provider to disclose clinical information (U98.0 and U98.1). ICD-10 codes, developed by the World Health Organization, are standard codes used to describe medical and health information and were introduced in the South African private health sector in 2005 [25]. Medical schemes utilize these standard codes for easy identification of diagnoses, especially those classified as PMBs and chronic, and appropriate reimbursements. The absence of or use of non-specific ICD-10 codes limits the ability to identify important patient health information to support public health research and reporting [26]. There may, therefore, be an underestimation of the prevalence of coexisting conditions in our study.

Secondly, our results showed that the majority (97.21%) of medicine items reimbursed with specific ICD-10 codes were indicated for acute conditions while only about 3% were for chronic conditions. This is supported by the higher proportion (82.10%) of all medicine claims reimbursed from patients’ acute benefits. This is to be expected because increasing age is a very important non-modifiable risk factor for chronic conditions [27,28]. Children are, therefore, less likely to develop chronic conditions. However, taking into account the potential complications of cancer and its treatment, chronic conditions may be identified in children on antineoplastic therapy. Chronic conditions identified in our study included hypertension which could be a result of the nephrotoxicity associated with the use of some chemotherapeutic agents such as cisplatin and ifosfamide [29]. Corticosteroids, which are mostly used as an adjunct therapy for some childhood cancers, have also been associated with hypertension in these patients [30]. Major Depressive Disorders (MDDs) and anxiety were also identified in our patient population and this could be attributed to psychological stress associated with cancer diagnosis and treatment [31]. Depression resulting from psychological stress is more prevalent in older children and adolescents with cancer due to the concurrence of the disease and their developmental stage [32]. This was confirmed in a study by Akimana et al. [33] which established that patients aged 10–17 years were four times more likely to be diagnosed with MDDs in comparison to younger children. This high prevalence of depression in older children and adolescents is confirmed by the prevalent use of antidepressants in the 10–14 and 15 < 19 years age group, compared to the other age groups in our study population.

Other chronic conditions identified in the study included epilepsy and asthma. The coexistence of epilepsy and cancer in our study is supported by the relatively high use of antiepileptic drugs as observed from our analysis based on the main pharmacological classes of all medicine claims. Epilepsy, which is mostly characterized by seizures, may be drug-induced, result from metastasis of primary brain tumor, or may be a complication of leukemia with brain involvement [34,35]. Epilepsy has been indicated as the most common chronic neurological condition in children [36] with a prevalence of 0.7% reported in rural South African children [37]. Some studies in South Africa have demonstrated a 9–34% prevalence of asthma in children in the general population, with a higher rate in residents of urban communities [38,39]. This confirms asthma as one of the prevalent chronic diseases in children. Asthma as a coexisting condition in our study population is supported by the prevalent use of anti-asthmatics, bronchodilators, antihistamines and corticosteroids. This is in contrast with previous studies that established a possible reduction in the prevalence of asthma symptoms and the need for asthma preventive therapies in children receiving chemotherapeutic agents [40,41]. It should, however, be noted that data for this study were expressed descriptively because of low numbers, and comparison with the prevalent use of asthma preventive therapy in patients unexposed to chemotherapeutic agents was outside the scope of this study.

Thirdly, diseases of the respiratory system were the most prevalent acute conditions overall, followed by diseases of the GIT, disorders of the skin, and diseases of the musculoskeletal system. Acute tonsillitis, acute upper respiratory tract infection, and acute bronchitis were the most common respiratory system diagnoses recorded on medicine claims. In support of these results, the DU90% analysis showed that antimicrobial agents, respiratory system agents, and analgesics, i.e. agents that are mainly used in the management of respiratory infections [42], were the three most prevalent pharmacological classes of medicines claimed with non-specific diagnostic codes. This may suggest a higher prevalence of respiratory diseases than was recorded in our patient population based on specific ICD-10 codes. Respiratory diseases are one of the leading causes of morbidity in children in the general population and represent approximately 25% of primary care consultations [43,44]. The immunosuppression associated with chemotherapeutic agents makes children undergoing cancer treatment more susceptible to respiratory diseases, notable among them being infections [45]. This is confirmed by results of previous studies that indicated the respiratory system as the common site of infections in children on antineoplastic therapy, with respiratory infections representing 16–23% of infectious episodes in these patients [46, 47]. These infections may be from bacterial, viral, or fungal origins [45]. Respiratory infections of bacterial origin, together with possible superinfection of respiratory viral infections with bacteria [48], is supported by the high prevalence of antimicrobial agents. Respiratory viral infections, especially influenza and respiratory syncytial virus infections which are common in immunocompromised children [49–52], are characterized by symptoms such as cold and cough [53,54]. This could also explain the prevalent use of medicines for coughs and colds (63.93%) among the respiratory system agents. The use of analgesics and antipyretics for the management of fever, a primary sign of infection resulting from neutropenia in children receiving chemotherapeutic agents [55,56], could explain the prevalent use of analgesics in this study.

Acute toxicities of the GIT, which include constipation, nausea, vomiting, diarrhea, and susceptibility to gastrointestinal infections, are often associated with cancer chemotherapy [57,58]. Chemotherapeutic agents such as cyclophosphamide are associated with mucosal ulceration, which predisposes patients to gastroenteritis [59]. Gastroenteritis is characterized by diarrhea, nausea, vomiting, and abdominal pain, and is an important cause of morbidity in children especially those who are immunocompromised [60–62]. Mucosal ulceration and its associated gastroenteritis from chemotherapeutic agents are supported by the high proportion of medicine claims associated with gastric ulcer (21.4%) and non-infective gastroenteritis (19.0%) among the diseases of the GIT in this study. This is also confirmed by the high prevalence of acid reducers among the GIT agents in our population.

The skin is prone to toxicities of chemotherapeutic agents since their mode of action involves targeting rapidly growing cells. Dermatological events, therefore, although rarely life-threatening, are usually reported in patients undergoing chemotherapy [63,64], with skin rashes, hyperpigmentation, and pruritus being the most prevalent condition. This could account for skin disorders being the third most prevalent disease group in our patient population.

The higher proportion of males as compared to females (68.79% vs. 31.21%) in the study population could account for the higher proportion of non-cytotoxic medicine claims in males. A comparison of the prevalence of coexisting conditions in males and females was, however, limited by the incomplete ICD-10 diagnostic codes on medicine claims. The highest proportion of non-cytotoxic medicine claims over the study period in the 5–9 years age group can be attributed to the majority of the study population (34.10%) falling within this age group. The majority of medications classified as CNS agents, musculoskeletal agents, genital system agents, dermatologicals, and urinary system agents, are mostly utilized in the adolescent age group. This may, therefore, account for the high prevalence of these pharmacological classes among the adolescent age group. For example, the high prevalence of acne in adolescents [65] makes the use of acne preparations and contraceptives relatively common in this age group. Contraceptives are also used in the management of menstrual disorders such as amenorrhea and menorrhagia in adolescents [66].

5. CONCLUSION

Most coexisting conditions in children and adolescents on cancer therapy in the section of the private health sector studied were acute conditions and included microbial infections and diseases of the respiratory system. Antimicrobial agents, respiratory system agents, analgesics, ENT and GIT agents were the top five most prevalent pharmacological classes of non-cytotoxic medications utilized by the study population. The high prevalent use of antimicrobial agents in this study indicates the need for the integration of antimicrobial surveillance programs into childhood and adolescent cancer care to curb antimicrobial infections.

CONFLICTS OF INTEREST

The authors declare they have no conflicts of interest.

AUTHORS’ CONTRIBUTION

The study was conceived by JRB, MNO, HS and MSL. MSL extracted and analyzed the medicine claims data. MNO interpreted findings and drafted the report under the supervision of JRB. All authors reviewed and approved the final version.

FUNDING

Financial support was received from the North-West University [30901979] and the National Research Foundation [grant number 118959]. These funding bodies were, however, not involved in the design of the study, analysis of the data, interpretation of the data, or the writing of the findings.

ACKNOWLEDGMENTS

We wish to thank the PBM company for allowing to use the database for the study, Ms Anne-Marie Bekker for her support with data extraction and analyses, and Mrs. Hoffman for help with the references. We acknowledge the North-West University and the National Research Foundation for providing financial support for this study.