Comparing the use of Arabic decision aid to usual care

Methods

A multi-centered randomized control trial (RCT) was used to evaluate the effect of an Arabic language decision aid (DA) created for metastatic colorectal cancer (mCRC) patients on their understanding of mCRC and anxiety levels. The study commenced in March 2016 and recruited its last patient in October 2018. Follow up was completed in April 2019.

Potential study participants were evaluated by the oncology clinic nurse through hospital referral or clinic appointment. Once a patient was identified as meeting the study inclusion criteria (seen below), they were asked to participate in the RCT. Written informed consent was provided by patient or their guardian; their chart was marked with a study label indicating that the patient was participating in a clinical trial. Each participant was allocated to a predetermined randomized arm based on the recruitment entry number.

All who met the following inclusion were recruited: age between 21 and 75 years of age; confirmed diagnosis of metastatic carcinoma of the colon or rectum; patient or companion must be able to read and write; written informed consent is obtained from each patient.

Patients were excluded from the study if they: were not a candidate for chemotherapy because of medical reasons; had previously received chemotherapy for mCRC (adjuvant chemotherapy permitted); were illiterate and/or cognitively impaired; had an uncontrolled psychiatric condition or any other condition that impairs ability to make decisions (namely dysphasia, severe comorbid illness, extreme anxiety, or distress, as assessed by the treating physician); had a life expectancy less than 6 weeks.

Patients were recruited from 4 main oncology centers in Saudi Arabia: King Fahad Medical City (KFMC), Riyadh; King Khalid University Hospital (KKUH), Riyadh; King Saud Medical City (KSMC), Riyadh; and King Fahd Specialist Hospital, Dammam (KFSH-D). The RCT attained ethical approval from all participating hospital Research Ethics Board (IRB) to introduce the DA tool that would be used in addition to the usual care.

The Ottawa Decision Support Framework was used as a reference11 to develop a video-based decision aid specifically for this randomized control trial in the Arabic Language. The framework of the Ottawa decision support tool lists the 2 treatment options (with or without chemotherapy), and it presents the treatment benefit, response rates, and estimated survival rates. The information was presented using graphic illustrations and numeric estimates. The tool was developed to complement usual care rather than replace the consultation provided by the physician.

The primary outcome for this study was to evaluate the patient’s understanding of mCRC prognosis and treatment information. Patient’s understanding was assessed using a questionnaire that covered the natural history of colorectal cancer, its recurrence, and the benefits, risks, and survival outcomes associated with chemotherapy by response options of ‘Yes, No, and I do not know’.12 The questionnaire was in Arabic and was piloted on 20 patients to ensure language comprehension before use. The scoring was based on the percentage of correct responses to determine the level of understanding on the impact and adverse effects of chemotherapy. Understanding was grouped into 4 domains: First, chemotherapy-induced toxicity which includes fever (more than 38 degrees) and severe diarrhea (more than 6 times above normal) with “Yes” as the correct answer. Painful numbness in the limbs and skin ulcers in the feet and hands (allergies) had “No” as the correct answer. The second outcome with chemotherapy includes chemotherapy helps people live longer and chemotherapy helps you overcome some of the symptoms and problems you experience as “Yes”. Third outcome without chemotherapy, which includes the following: my condition will worsen without chemotherapy as “Yes”. Finally, the goal was to understand that the disease is incurable; chemotherapy helps full recovery from cancer as “No” The percentage of correct answers was calculated as the number of correct answers in the domain by the number of all answers in the same domain.

The assessment of patient understanding was measured at baseline and included a repeat measurement at one month to assess how much information patients had retained. Patient understanding at baseline was collected from the total RCT sample population; however, the follow-up at one month was a partial sample of the total due to a misinterpretation repeated outcome measurement collection. As such, a correction to collect one month was initiated once identified. While a repeated measure at 6 months was viewed to reflect experience rather than understanding.

The second outcome measure was patient anxiety levels. For patient anxiety, the general anxiety disorder-7 (GAD7) was chosen after extensive review of available literature.13 The tool has been validated for use in both the English and Arabic language.14 General anxiety disorder-7 is the most common version and includes 7 items for assessing anxiety disorder. All items are rated on a 4-point scale 0-3 (namely “not at all”, “several days”, “more than half the days” and “nearly every day”). The higher scores are an indication of greater anxiety. The GAD7 is user-friendly for those with at least a sixth-grade reading level, and the overall Cronbach’s alpha was 0.91. The total score for the 7 items range from 0 to 21. These are split into levels of anxiety.15 The levels of anxiety are split into 4 groups: i) 0-4 minimal anxiety; ii) 5-9 mild anxiety; iii) 10-14 moderate anxiety; iv) 15-21 severe anxiety.

Each study participants provided their sociodemographic data and disease characteristics at baseline along with evaluation of anxiety level through GAD7. This took place immediately after their consultation with the physician to assess their understanding and anxiety level. At this point, the study participant reached their initial decision toward a treatment option. At one month and then 6 months’ post-treatment decision, a repeated measure was collected regarding anxiety levels of study participants.

Power calculations indicated that 90 patients were needed. The outcome variable for anxiety level was measured by GAD7 with a seven-item range score from not at all as 0 and nearly every day as 3, with minimum score of 0 and a maximum of 21. At least 45 subjects were needed to detect a difference in the mean GAD7 score between decision aid and usual care of at least 3 points with a standard deviation of 5 at 80% power and 5% alpha. This calculation used Pi face.

The allocation of each study participant was through a computer-generated stratified randomization lists as a case (DA with usual care) or control arm (usual care alone) for each participating hospital. The RCT design did not require the health care providers and participants to be blinded due to the nature of the study; the DA is administered as an intervention in the form of video-based informational tool. The DA tool provided a coherent description of prognosis, diagnosis and treatment options as covered in usual care by an oncologist. The oncologists were informed of patient allocation of randomization arm only after the patient received the DA or usual care. Patients allocated to the DA arm were counseled not to share it with others in the waiting room to avoid contamination of information with those in the control arm.