Minimising treatment-associated risks in systemic cancer therapy

In order to illustrate the most common ADRs in oncology patients in a cancer centre in Australia were observed and the incidence as well as the predictability, preventability, and severity of the occurred ADRs were assessed. Among the ten most common ADRs constipation ranked first but was connected directly to the use of opioids rather than to cytotoxic chemotherapy. Nausea and vomiting, fatigue, alopecia, drowsiness and myelosuppression ranked second to sixth. Of these ADRs 88% were predictable and about 50% even probably preventable, because of inadequate use of preventative measures [4].

As well as the incidence of ADRs in modern chemotherapy the patients’ perceptions of the impact of these ADRs on well being and quality of life are increasingly taken into account. Patient perceptions have changed markedly over the last two decades. Whereas in 1983 physical symptoms such as nausea, vomiting and hair loss were most troubling from the patients’ point of view, in 2002 psychosocial complaints ranked among the top ten symptoms, with the complaint “affects my family or partner” rated as the most severe ADR. Alopecia and fatigue ranked second and third place [5, 6].

As a consequence, the management of ADRs has become a focus of clinical research and new drugs such as aprepitant, a neurokinin-1-receptor antagonist for the prevention of nausea and vomiting, were developed [7]. Despite these new therapeutic options the minimisation of toxic effects of chemotherapy still seems to be a challenging task.

The first and most important step to a better management of ADRs is the formulation and implementation of evidence-based clinical practice guidelines for the different symptoms in a multidisciplinary approach. Several studies have shown a positive effect of guidelines for antiemetic prophylaxis and therapy on both clinical and economic outcomes [8‐11]. The most important influencing factors seem to be the appropriate use of 5-HT₃ receptor antagonists and the application of oral drug formulations. Nevertheless, altering the usage patterns of physicians to comply with evidence-based guidelines seems to be difficult. Methods such as the use of computerised decision support systems and educational outreach mechanisms appear to be most effective particularly when used in combination [12]. The pharmacist can support and promote the adherence to guidelines [13].

A second step to a better management of ADRs is the implementation of standardised chemotherapy order forms that also contain supportive care medication. By using standardised order forms a more appropriate prescribing of antiemetics based on the level of emetogenicity of administered chemotherapy and thus a reduction in drug expenditure can be achieved [14]. Such order forms should be elaborated following a multidisciplinary approach including physicians, pharmacists and nurses.

The large number of prescribed drugs administered to cancer patients leads to a high potential for drug–drug interactions. As indicated previously many cancer patients also use over-the-counter medication as well as alternative and complementary treatment options. Frequently, cancer patients suffer from concomitant chronic diseases that require the intake of other drugs, further increasing the risk of interactions. This problem is often underestimated in oncology as highlighted in several review articles [15‐18]. Moreover, drug–food, drug–disease and drug–diagnostics interactions have to be considered that are, however, beyond the scope of this article.

Drug–drug interactions must be avoided as they can lead to overdosing or underdosing of anticancer drugs with the consequence of toxicity or a loss of effectiveness, respectively. However, not all published interactions are also clinically relevant. A rational and safe drug therapy includes a check for potential drug–drug interactions based on the individual medication and a decision whether and how identified drug–drug interactions have to be considered or not.

Knowledge of the mechanisms of interactions is crucial to assess the clinical relevance of an interaction. Drug–drug interactions are usually classified as pharmaceutical, pharmacokinetic or pharmacodynamic interactions (Table 2). The risk of interactions can differ substantially between individual drugs, even within the same class of drugs. This can be considered when selecting a particular drug for an individual patient.

For some drugs with a high potential for interactions, therapeutic drug monitoring can be useful to detect and control interactions. The advantage of this approach is that the dosage of the drug of risk can be adapted to the measured individual plasma concentrations. Thus the treatment of both interacting drugs may be continued. Examples are anticonvulsant, antidepressant and antifungal drugs which are often administered to cancer patients as supportive medication [19].

In general, it is almost impossible for the physician to keep in mind all of the interaction mechanisms and potential consequences for the individual patient. Regular interaction checks by the pharmacist may solve this problem as he is the only health care professional who may have an overview of the drugs prescribed by various physicians and the medication taken by the patient on his own initiative. The avoidance of drug–drug interactions should be regarded as a multidisciplinary task [20].

The basis for each interaction check is to take the medication history of each patient and to update it regularly. Nowadays software tools are available facilitating rapid interaction checks and providing information on the mechanism and clinical relevance of an interaction. Each hospital and community pharmacy should have access to at least one of these tools.

In addition to the DRP described above, there are numerous risks for the occurrence of medication errors along the therapeutic path in oncology. A medication error is defined as any preventable event that may cause or lead to inappropriate medication use or patient harm, while the medication is in the control of the health care professional, patient, or consumer [21]. In 2000 the American Institute of Medicine published their report “To err is human: Building a Safer Health System” which analysed flaws in the health system and offered suggestions for its improvement [22]. Based on the results of a study conducted in Utah and Colorado in 1992 an incidence of about 3% adverse events in hospitalised patients was found, of which the majority of the non-operative events were adverse drug events (ADE) [23]. The authors concluded that iatrogenic injury continues to be a significant public health problem. Due to the narrow therapeutic range of many anticancer drugs, the impact of an ADE is more serious (and at worst lethal) compared to other systemically administered drugs. Antineoplastic agents were found to be the second most common group of drugs which caused lethal medication errors [24]. As well as quality and safety issues, the reduction of medication errors has a financial effect.

The first source of medication errors is the prescription and ordering process: the wrong protocol may be chosen, the cumulative dose may be mixed up with the single dose, the route of administration may not be clearly indicated or misinterpretation of the physicians’ handwriting may lead to errors. A study carried out in a cytotoxic preparation unit in a French hospital revealed in only a 6 month period more than 300 medication errors out of 1,262 prescriptions for 285 patients. Most errors (>70%) were found to be simple physicochemical incompatibilities, e.g. incompatibility of the drug with the used matrix. However, also over- and underdosage as well as wrong medications were identified [25]. The authors conclude that most of these errors could possibly be avoided by a computerised prescription network. Computer-assisted solutions enable physicians to order the medication electronically which is nowadays one way to reduce the incidence of those errors. Mekhjian et al. evaluated the benefits of physician order entry systems and found that transcription errors could be reduced and speed in the ordering process could be improved. Even the duration of the stay in hospital could be reduced by the system [26]. Particularly in oncology, the computerised physician order entry (CPOE) has obvious advantages. A recent study has shown that CPOE in electronic medical records improves completeness of the medical record compared to paper charts independent of regimen complexity [27]. Prepared protocols facilitate correct ordering and double checking by the programme itself and by the pharmacist responsible. In addition, supportive medication, such as hydration or the antiemetic or antiallergic medication can be automatically proposed depending on the selected regimen. Moreover, the risk of misinterpretation of handwriting is eliminated. Recently, a CPOE system with integrated clinical decision support has been implemented in Ontario (Canada). This system, for example, alerts the clinician when maximum cumulative doses are reached, calculates body surface area, and estimates the creatinine clearance of the patient [28]. With all the advantages of these systems it has to be taken into account that new risks may arise from those new processes such as information errors due to the lack of clearness on the monitor [29].

The next step after ordering, the compounding of cytotoxic drugs also holds various risks, especially when performed on the ward. The implementation of central cytotoxic preparation units in pharmacy departments in the late 1980s has been one of the first measures to standardise the process in order to increase safety. Nevertheless, of 30,819 preparations surveyed in a study on incidence and risk factors of preparation errors 140 were found to be defective (0.45%) [30]. Less than half of those cases were classified as major errors including wrong dosage, wrong labelling or the use of incompatible diluents. Although the incidence is fairly low it must still be the aim to reduce this further by analysing the risk factors such as workload. Centralisation itself offers many more options for safety improvement. Standardised pharmaceutical validation leads to an increased interception of medication errors in antineoplastic treatment. This can be explained by the improved knowledge of the pharmacists involved in the validation process. In consequence of a standardised pharmaceutical validation process in the hematology–oncology department of a 550-bed university hospital the number of detected medication errors that did not reach the patients could be increased by 41% (from 14.08 medication errors/1,000 patient days to 19.83 medication errors/1,000 patient days) in the second year after the introduction of pharmacists in the multidisciplinary oncology team [31]. Using their expertise, pharmacists can facilitate protocol development and adherence, dose verifications and education of other health care practitioners [32].

Finally, storage and administration errors of cytotoxic drugs form a group of major medication errors with serious consequences for the patient. Therefore, it is mandatory that the assigned staff are well trained in all aspects of chemotherapy storage and administration. For many drugs, special storage conditions, e.g. refrigeration or protection from light, have to be considered. Detailed knowledge of the administered drugs is also very important in order to educate the patient and to react appropriately when incidents such as extravasations occur. In case of extravasation the necessary measures have to be initiated. Different drugs require different measures which the oncologically trained staff needs to know. An extravasation kit including essential items for first aid must be made available on site [33]. Incorrect or delayed treatment can cause serious damage to the patient, resulting in even the loss of a limb.

Various surveys have been undertaken in order to find solutions for the prevention of medication errors. The non-punitive reporting of errors is a necessary requirement to be able to detect, analyse and consequently minimise the incidence and severity of medication errors in oncology. Safety measures with error reporting and analysis have been developed and implemented in order to improve the entire system [34]. The most effective improvements of treatment performance have been achieved by multidisciplinary system approaches integrating physicians, pharmacists and nurses [2, 32, 34‐36].

Adherence (or compliance) is generally defined as the extent to which a person’s behaviour, taking medication, following a diet, and/or executing lifestyle changes, corresponds with agreed recommendations from a health care provider [37]. Non-adherence strongly compromises the success of a patient’s therapy, results in additional, potentially unnecessary, diagnostic and therapeutic procedures and thus generates further costs and possibly health problems as a consequence of the treatments themselves. It is important to note that the traditional concept that patients are solely responsible for taking their medication (and therefore non-adherence is a patient-driven problem) is misleading. The WHO understands adherence as a phenomenon consisting of five dimensions (Table 3). The magnitude of this problem is further emphasised by the fact that adherence rates for many long-term drug therapies range from only 40 to 78% [38, 39].

It is a difficult task to assess patient adherence reliably. Patient self-reports, rates of prescription refills, patient diaries or pill counts as the sole basis for the measurement of adherence have been shown to be inadequate [38‐40]. It has been demonstrated that these methods overestimate the degree to which patients adhere to their tamoxifen regimen [40]. The use of microelectronic adherence monitoring (Medication Event Monitoring System, MEMS^TM, Fig. 1) provides a valuable estimate of the timing of events and insights into patients’ behaviour in taking medication. However, to prove actual intake of the medication, additional plasma concentration monitoring would be necessary. To obtain an optimal measurement of patient adherence, a combination of several methods is suggested [38].

Studies conducted in various disease areas have shown that a wide variety of methods to improve adherence is available. A recently published review classifies these methods into four general categories: patient education, improved dosage, increased hours when clinics are open, and improved communication between health care providers and patients [38]. All four categories require a collaborative approach including all members of the multidisciplinary health care team. Due to their position in the chain of health care providers, this requires the contribution of pharmacists, not only in the hospital setting but also in community pharmacies [41]. It has been shown that no single intervention strategy appears consistently more effective than another. Successful methods are complex and comprehensive as well as labour-intensive [38, 42]. The Cochrane Review on Interventions to Enhance Medication Adherence concludes that there is no evidence that low adherence can be “cured” and hence efforts to improve adherence must be maintained for as long as the treatment is needed [43].

Cancer patients seem to benefit especially from interventions towards an optimised adherence, resulting in improved outcomes [42]. Research investigating adherence in oncology settings has been mostly focussed on palliative care and supportive medication because chemotherapy has mainly been administered intravenously in hospitals or clinics. However, with the increasing availability and importance of oral anticancer drugs such as capecitabine, etoposide, vinorelbine, erlotinib and sorafenib, patients increasingly take over the responsibility for the correct administration of their prescribed therapy. Although cancer patients might be more adherent than other patient groups due to a high level of motivation, the slightest non-adherence can endanger the therapeutic goals [39].

Only few studies have been published investigating the level of adherence of cancer patients. In a cohort of outpatients receiving chemotherapy for haematologic malignancies adherence to the allopurinol and prednisolone prescription was measured based on plasma concentrations. It was shown that control patients without intervention were adherent only 17% of the time with allopurinol and 27% of the time with prednisone. Three specially developed intervention packages including education and home visits resulted in increased adherence rates of 44–48% and 33–38% of the time for allopurinol and prednisone, respectively [44]. Another study assessed adherence in patients with breast cancer receiving oral cyclophosphamide and found that 43% of the patient population met criteria for non-adherence according to both behavioural and dosage definitions [45]. In a recently published study in a cohort of 2816 women adherence to tamoxifen treatment was investigated using prescription refill data: 22.1% of the patients exhibited a discontinuation of treatment (non-persistence) within 1 year [46]. An open controlled trial is currently undertaken at the University of Bonn investigating the impact of intensified pharmaceutical care provision on the adherence of patients receiving oral capecitabine using the MEMS™-technology. Pharmacists’ interventions include the extensive provision of drug information to the patients and health care professionals, routine checks for drug interactions, preparation of written drug intake plans for patients, regular pharmacist consultations and structured documentation of the pharmaceutical care process [47].