The Rationale of Patients with Early-Stage Papillary Thyroid Cancer for Accepting or Rejecting Radioactive Iodine Remnant Ablation

Dear Editor:

For patients with early-stage (low-risk) papillary thyroid cancer who have had a total thyroidectomy, selective use of radioactive iodine remnant ablation (RRA) has been recommended by the American Thyroid Association, since it may (i) facilitate initial staging (by post-therapy whole-body scan), (ii) facilitate disease surveillance (based on serum thyroglobulin measurement or whole-body radioactive iodine scan) due to ablation of a thyroid remnant, and (iii) provide a potential theoretical tumoricidal effect on persistent occult thyroid cancer cells, although data are conflicting on its effect on disease recurrence (1). There is significant variability in the utilization of RRA in the United States, and most of this variability cannot be explained by clinical or pathologic characteristics of patients (2). Patients' rationale for accepting or declining RRA is an important, relatively unstudied variable, for consideration in understanding the heterogeneity of use of this intervention.

We recently conducted a randomized controlled trial testing a patient-directed computerized decision aid (DA), on the topic of choice to accept or decline RRA in early-stage papillary thyroid cancer (3,4). The details of the trial were recently published (3,4); 37 consenting patients were randomized to a one-time viewing of the DA Web program (in addition to usual counseling/care from their treating physicians), and 37 consenting patients were randomized to usual counseling/care alone (controls) (3,4). The DA significantly increased patients' medical knowledge and reduced their decisional conflict, but did not appear to impact the RRA utilization rate (4). This trial was registered at ClinicalTrials.gov (NCT01083550) and approved by the University Health Network Research Ethics Board (3,4).

Approximately 6 months after randomization, we contacted all participants by telephone to ascertain whether RRA was received or not (with no intention to receive it), and we asked, using an open-ended question, for an explanation of the rationale for the decision. If the choice was not yet finalized, we recontacted the patient several months later. The reason for contacting patients beginning 6 months after randomization was to ensure that the decision was recent enough for patients to recall the rationale. In keeping with a concurrent mixed-methods research design approach (5), the following steps were taken in data collection and analysis of the decision rationale data: (i) the narrative response data were collected in detailed notes by the research staff member; (ii) responses were coded by two independent reviewers and consensus achieved on coding; (iii) theme categories of related categories of codes were created by a coder as well as an investigator with clinical content expertise (A.M.S.), and consensus was achieved on final themes; and (iv) the number of mentions of each theme category was ascertained (frequency-scale coding).

Data collection on treatment choice and rationale was completed for all 74 patients in the trial. RRA was taken by 23.4% of patients [18/74, including 11 in the DA group and 7 controls (4)]. In response to our question on patients' rationale for or against RRA, we collected a total of 159 statements from the study population of 74 patients (including 83 statements on source of rationale information and 76 statements on the rationale itself [with some patients citing more than one source or rationale]). By far, the treating physician was the most commonly cited source of information relating to treatment rationale (51 statements—21 DA group, 30 control group), and this was followed by the personal research or the internet (20 statements—equally divided between DA and control groups), the DA under study (11 statements—DA group), and a family member who was a physician (1 statement—DA group). No reason for the choice, other than the information source, was reported by 21 individuals (9 DA group, 12 control group). For the 18 individuals treated with RRA, the rationale included the concern about the thyroid cancer (e.g., possible disease persistence or future recurrence; 7 statements—5 DA group, 2 control group), destruction of remnant thyroid cells and facilitating disease surveillance (6 statements—4 DA group, 2 control group), and weighing multiple issues (such as possible risks and benefits of the options) in the context of personal values (3 statements—1 DA group, 2 control group). For the 56 individuals who did not receive RRA, the reason for rationale included a perception of early-stage disease and good prognosis (34 statements—16 DA group, 18 control group), potential negative aspects of RRA (including possible short- or long-term side effects or aspects of the procedure itself such as thyroid hormone withdrawal [in the event of lack of availability of recombinant human thyrotropin] or post-treatment isolation; 17 statements—11 DA group, 6 control group), insufficient evidence of RRA treatment benefit (4 statements—1 DA group, 3 control group), weighing multiple issues (such as possible risks and benefits of the options) in the context of personal values (4 statements—1 DA group, 3 control group), and pregnancy (1 statement—control group).

In summary, physicians were the most commonly identified source of medical information upon which the decision to accept or reject RRA was based, and patients' rationale for the choice was largely explained by perceptions of the degree of threat of the malignancy (present or future), relative to the perceived importance of negative aspects of the treatment (including uncertainty of treatment benefit). In our study, standardized presentation of information in a DA did not appear to dramatically alter the utilization of RRA (4), nor the types of reasons that were used to explain the treatment choices.

Some strengths of this study include the systematic approach to collection and analysis of data within a randomized controlled trial and the lack of any loss to follow-up of study participants. Some limitations of the study include a lack of audio recording of phone interviews, a lack of objective data on the physician counseling encounter (such as videotapes or audio recordings), and recruitment of study participants from a relatively localized geographic area (Ontario, Canada).

These data provide some insight into the current heterogeneity of utilization of RRA in early-stage papillary thyroid cancer in North America, particularly relating to variability of patient perspectives. It is important to note that an increasing proportion of cancer patients wants to be involved in medical decision making (6), so it is highly relevant for clinicians to be aware of patients' rationale for treatment choices, and clarify any potential misperceptions, if present.