Usefulness of Cellular Text Messaging for Improving Adherence Among Adolescents and Young Adults with Systemic Lupus Erythematosus

Patient population. Visit adherence population

With approval of the institutional review board, all patients diagnosed with cSLE⁴ and followed in a pediatric rheumatology clinic were identified using a lupus registry. For inclusion, patients had to be between the ages of 13 and 25 years and have unlimited access to cellular text messaging.

Medication adherence population

A subset of patients who were part of the visit adherence population were recruited for participation in a substudy to assess the benefits of CTMR on adherence to medication. To participate in the medication adherence substudy, patients had to be at least 15 years of age, be treated with HCQ for cSLE, and have a minimum disease duration of 6 months. Younger patients were excluded because they were unlikely to be responsible for taking their medications on their own. Patients with active neuropsychiatric SLE symptoms were excluded, as were those with other chronic diseases (e.g., diabetes mellitus) that might influence medication use.

Study questionnaires and disease measures

Patient demographic data including age, sex, race, insurance status, estimated socioeconomic status, and distance of residence to the medical center were noted. The latter was measured using Google Maps directions tool for each patient’s postal code in relationship to the hospital address. Estimated median family incomes were identified using Geocoding (Federal Financial Institutions Examination Council, Fairfax, VA, USA; Website: http://www.ffiec.gov/Geocode/default.aspx) to provide an overall economic status for each patient. Information regarding disease damage was gathered at the time of each medication adherence visit and was measured by the Systemic Lupus International Collaborating Clinics/American College of Rheumatology Damage Index (SDI)⁵, the Systemic Lupus Erythematosus Disease Activity Index (SLEDAI), and the physician assessment of disease activity (by visual analog scale; range 0–10, 0 = inactive disease, 10 = very active disease)⁶. The number of emergency room (ER) visits and the number of hospital admissions were monitored among patients participating in clinic visit adherence.

Study interventions

Cellular phone number and cellular service carrier were obtained from patients at the beginning of the study and monitored periodically for changes throughout the intervention period. CTMR were sent over a free e-mail network system (e-mail to text).

Visit adherence intervention

A CTMR to the patient’s cellular phone was sent 7 days, 3 days, and 1 day prior to each scheduled followup clinic appointment. CTMR were individualized for each patient by including the scheduled time of the upcoming clinic appointment (Table 1). Additionally, CTMR were sent in the event that a patient failed to schedule a followup appointment within the first 2–3 weeks after the preceding clinic visit. For each patient, clinic attendance data were compared with his/her visit adherence between January 1, 2006, and June 30, 2008, prior to the CTMR intervention, as well as his/her visit adherence after the CTMR was discontinued from November 15, 2009, through June 15, 2010.

Medication adherence intervention

Patients participating in the medication adherence substudy were randomized 1:1 to either standard of care (SOC group) or the CTMR intervention (CTMR group). Participants were followed for up to 6 study visits that occurred during regular clinic visits 2–4 months apart for up to 14 months. The SOC group received printed information about the benefits and side effects of HCQ as per current standards of clinical care. The CTMR group received a standardized daily CTMR for HCQ intake as prescribed (e.g., once or twice per day) in addition to the printed information sheet that was given to the SOC group. A CTMR was sent to all patients at a set time of day (e.g., morning or evening), according to self-report of HCQ intake (Table 1). Changes to dosing of HCQ were noted following each visit, and CTMR were altered accordingly.

Measurement of adherence. Visit adherence

Using electronic scheduling databases and information prospectively recorded in the electronic medical record, the following visit-specific information was collected for each patient in the visit adherence cohort: total number of rheumatology clinic visits attended, total number of “no-shows” (i.e., clinic appointment scheduled but never attended or canceled), total number of cancellations (i.e., clinic appointment scheduled and canceled by the patient in advance of the day of visit), and the suggested time interval between clinic visits as per the treating pediatric rheumatologist.

Adherence to clinic visits was measured as the percentage of clinic visits that occurred “on time” (i.e., within the timeframe recommended by healthcare providers). Visit attendance was deemed “on-time” when patients were seen in the rheumatology clinic within 1 week of a 1-month recommended followup time period, 2 weeks within a 2-month followup, 3 weeks for a 3-month followup, etc. Acceptable clinic visit attendance rates were defined as ≥ 80% adherence as per the healthcare provider visit recommendations⁷.

Medication adherence

Due to its beneficial effects, most patients participating in the lupus registry are treated with HCQ for cSLE⁸. The half-life of HCQ is around 40 days, and drug levels can be measured in the blood, reflecting longer-term exposure to HCQ⁹. As there is no criterion standard for measuring medication adherence in a clinical setting, adherence to HCQ use was evaluated using 3 different approaches: (1) patient’s self-report on the validated Medication Adherence Self-Report Inventory (MASRI)¹⁰ at each study visit; (2) whole-blood levels of HCQ were measured by a commercial laboratory (NMS Labs Inc., Willow Grove, PA, USA) with methods as described¹¹; and (3) pharmacy refill adherence, defined as the percentage of the number of HCQ doses dispensed (numerator) and the number of doses prescribed (denominator) for the period of time between study visits and pharmacy refill dates, with pharmacy refill information serving as primary measure of medication adherence⁷. These approaches were measured at baseline and at all followup visits (about every 2–4 months).

Patients with HCQ adherence > 80% were considered sufficiently adherent, while those with < 80% adherence were considered nonadherent to HCQ as per pharmacy refill information⁷. HCQ adherence was measured starting 9 months prior to CTMR and continued until 7 months after the CTMR for HCQ were discontinued.

Statistical analysis

Descriptive analysis included means and SD for numerical data and percentages for categorical data. Associations between patient demographics and visit adherence were determined in univariate analysis to identify confounders of nonadherence using contingency tables or logistic regression analysis for categorical data.

The effects of CTMR on visit adherence were assessed by comparing patient adherence behavior over time: preintervention (pre-CTMR), during intervention (CTMR period), and postintervention (post-CTMR). Similarly, HCQ adherence was assessed over time, with consideration of whether the patient received CTMR.

Paired t tests (for visit adherence), 2-sample t tests with unequal variances (for HCQ adherence), and mixed models were carried out to assess the effects of CTMR on visit and HCQ, respectively. Effect sizes (Cohen’s d) were calculated. By convention, t test effect size (d) values of 0.2, 0.5, and 0.8 are considered small, medium, and large, respectively. P values < 0.05 were considered statistically significant.