Engaging Patients in Clinical Trial Planning: Exploring the Concept of a Minimal Clinically Important Difference Among Patients with Obstetrical Antiphospholipid Syndrome

INTRODUCTION: Women with antiphospholipid syndrome are at increased risk of pregnancy loss. A large randomized controlled trial is still needed to confirm or refute the efficacy of low-molecular-weight heparin (LMWH) and aspirin (ASA) prophylaxis during pregnancy in women with antiphospholipid syndrome (APS) and past pregnancy loss. The minimal clinically important difference (MCID) is a statistical concept that quantifies the smallest amount of benefit from a treatment that a patient would identify as important, and is used to adequately power clinical trials. When planning a clinical trial, the MCID is most often determined by specialists in the field. How patients make sense of MCID as a concept, and how patients' perceptions of MCID compare to physician responses is unknown. In a challenging and high stakes situation, understanding different patients' perspectives must be considered when planning a large clinical trial.

METHODS: Consecutive patients were approached from a specialty Thrombosis clinic and from the TIPPS (Thrombophilia in Pregnancy Prophylaxis Study) screening logs between January 2017 and March 2018 in Ottawa, Canada to complete a survey and interview. Patients were eligible if they met the revised Sapporo/Sydney laboratory criteria for antiphospholipid syndrome and had at least 1 late pregnancy loss or 2 early losses (<10 weeks gestation). A similar survey was distributed to 350 specialists identified through 5 professional organizations between May-Sept 2016. Both patient and physician groups had access to an introductory video. The mean MCID between the patient group and physician group was compared using an independent t-test, with p<0.05 being significant. Patient interview data were collected and analysed iteratively, in keeping with constructivist grounded theory methodology.

RESULTS: There were 22 (33%) patients who completed surveys and 10 patients who completed semi-structured interviews. Among the physician group, 72 (20.6%) responded to the survey, including obstetricians (34.7%), hematologists (26.4%), general internists (18%), maternal fetal medicine specialists (6.9%) and other specialties (12.6%). Assuming a live birth rate of 60% in the ASA alone arm, the mean MCID answered by physicians was 11.6% (range 5-25%, SD 5.2), compared to the mean MCID answered by patients was 10.0% (range 5-20%, SD 6.4) (p=0.28). The mean number needed to treat (NNT) answered by both physicians and patients was 16. In the majority of cases from both groups, the related concepts of MCID and NNT did not match statistically (i.e. an MCID of 15% is a NNT of 7).

An overarching theme from the patient survey comments and interviews was that statistics were somewhat irrelevant. The majority of women thought that "any increase in a chance of live birth" was worth it. However, there was variation in patient responses where some women would only accept a certain success rate of LMWH to justify its use because of drug cost or side effects. Some women acknowledged that they would not use LMWH themselves, but still advocated for other patients to have the option of LMWH use for a chance of a live birth.

CONCLUSION: The mean MCID and NNT were similar between patient and physician groups. The low mean MCID and NNT values likely reflect the importance of the outcome studied (avoiding pregnancy loss) and the perceived safety of the proposed interventions of LMWH and/or ASA, which is supported by qualitative data from patient interviews. While the MCID and NNT responses were similar among physicians and patients, the understanding of statistics was somewhat limited in both groups. By better understanding patients' rationale and beliefs about LMWH use, we can better engage them in clinical trial planning and patient care and ensure that we are performing research that is relevant to patients.