While randomized clinical trials (RCTs) remain the gold standard in clinical research, they pose challenges for cannabinoid therapies due to the vast variability in cannabis formulations and the genetically influenced differences in individual responses. These complexities make standardization of dosages and compositions difficult and highlight the value of real-world evidence (RWE) during this still exploratory stage of cannabinoid research. RWE studies offer broader inclusion criteria, flexible dosing, and the ability to address syndromic conditions and comorbidities beyond narrowly defined core symptoms (Baumfeld Andre et al., 2020; Banerjee et al., 2022). They also enhance ecological validity and may detect long-term or rare side effects, which are often missed in controlled trials.

In our study, the real-life approach—employing personalized FCE dosages across diverse CPS conditions, with long-term use and psychosocial evaluation—was essential to capture therapeutic outcomes realistically. However, inherent limitations must be acknowledged (Crisafulli et al., 2023). Patients who experienced positive outcomes may have been more likely to complete the retrospective, cross-sectional online survey, introducing possible selection bias. Additionally, the absence of direct researcher-participant interaction during the survey process, while minimizing interviewer influence, meant that the survey data were based solely on self-reported subjective patient perception (except for side effects and other medications, which were complemented with information from clinical records).

Ultimately, pain, lack of libido, fatigue, and mood disorders, for instance, are internal, inherently subjective experiences, making them largely inaccessible to direct quantitative measurement—especially in their chronic forms. Therefore, self-reported outcome instruments are essential tools for assessing the efficacy of treatments for CPS and related mood or functional disorders, as they provide unique access to the patient’s pain experience and overall wellbeing (Mosko et al., 1989; Schierhout and Myers, 1996; Bovier et al., 2004; Cuijpers et al., 2010; Morso et al., 2011; Karibe et al., 2014; Altenstein-Yamanaka et al., 2017; Ohno et al., 2017; Murphy et al., 2018; Ben Tekaya et al., 2019; Vincent et al., 2019; Riddle and Reza Jafarzadeh, 2022; Turnbull et al., 2022; Eiger et al., 2024; Gupta et al., 2024; Huang et al., 2025). In fact, self-report measures are often regarded as the “gold standard” for evaluating pain-related outcomes, particularly when supplemented with objective indicators such as changes in the use of rescue treatments (i.e., additional medications or interventions) (Dworkin et al., 2005), as was done in this report.

Despite their strengths, self-reported outcome instruments pose notable methodological challenges. On the positive side, they are easy to administer, cost-effective, and capable of capturing nuanced, individual variations in symptoms that may escape clinical observation—provided they are carefully designed and clearly communicated to ensure patient comprehension (Turk et al., 2003; Turk et al., 2008; Patrick et al., 2011b; a). In the context of chronic pain research, such tools offer direct insights into pain intensity, affective distress, and the psychosocial and functional impacts on daily life—dimensions that are not adequately captured by physiological markers alone (Dworkin et al., 2005; Turk et al., 2016). Likewise, mood disorders such as depression and anxiety often manifest internally, making self-assessment scales crucial for diagnosis and monitoring (Kroenke et al., 2001).

In this study, we adapted elements from multiple instruments to develop a customized, multifactorial, yet simplified, Patient-Reported Outcome Survey. This tool was designed to allow patients to easily and simultaneously evaluate multiple aspects of their CPS using a unified, non-numerical scale (later translated into numeric scores for analysis). However, like all self-report tools—particularly those used in open-label or non-blinded designs, which is the case here—our instrument is susceptible to several sources of response bias. These include social desirability, recall inaccuracies, participant expectations, placebo effects, and behavioral changes stemming from participants’ awareness of being observed, i.e., the Hawthorne effect (Enck et al., 2013; Hrobjartsson et al., 2014; Shiozawa et al., 2014; Chiarotto et al., 2019; Wang et al., 2023). In our study, we relied on three key aspects to help mitigate these potential sources of bias: 1) the duration of treatment, which exceeded 6 months for most participants, likely reducing the influence of short-term placebo responses and patient expectations, as well as the impact of possible Hawthorne effect; 2) the observation of reduction in the use of other medications—both for pain and mood disorders—which offer supportive and objective evidence of treatment impact, and 3) the use of an online survey format, completed without any direct contact with physicians or researchers, which may have helped reduce social desirability bias as well as the Hawthorne effect.

The original treatment protocol was not designed as a research intervention, but rather as compassionate clinical care conducted by the physician. Lacking a standardized baseline, we developed a custom cross-sectional outcome survey to assess psychological, functional, and symptomatic aspects of CPS using a single response scale. Although this limits direct numerical comparisons with other studies, the use of a narrow scale enhanced comparability across patients and better captured the complexity of chronic pain syndromes as experienced individually. It also helped mitigate the limitations inherent in assigning numerical values to subjective perceptions—an issue particularly problematic given the significant influence of emotional states and hormonal levels on pain perception (Wijnhoven et al., 2006; Reicherts et al., 2013; Lumley et al., 2021). Thus, the focus of this study was not statistical generalization, but rather an in-depth understanding of patient-reported outcomes across multiple CPS presentations.

We also developed our own questionnaire to assess adverse effects associated with cannabis extract use because conventional instruments—such as the Common Terminology Criteria for Adverse Events (CTCAE), the UKU Side Effect Rating Scale, and the Systematic Assessment for Treatment Emergent Events (SAFTEE)—do not specifically address some acute or subjective effects commonly associated with cannabis, including red eyes, and the sensation of feeling high (Levine and Schooler, 1986; Lingjaerde et al., 1987; Trotti et al., 2003). Although often mild, these effects are clinically relevant when evaluating the tolerability of cannabinoid-based therapies. By including cannabis-specific symptoms, we aimed to enhance the ecological validity of our study and provide a more comprehensive picture of real-world patient experiences.

Furthermore, our questionnaire captured not only the presence of adverse effects but also their intensity (mild, moderate, or severe) and temporality (transient vs. persistent). This level of detail is typically absent from standard instruments, which tend to focus on general or class-specific events and rarely incorporate patient-centered measures of symptom duration or subjective severity. We believe our approach yields a more nuanced and clinically meaningful profile of tolerability, supporting better-informed therapeutic decisions in real-life settings.

Finally, our Patient-Reported Outcome Survey has not yet undergone formal psychometric validation for use as a generalizable statistical instrument to support comparisons across different studies. This represents an important next step in our research agenda, but it falls beyond the scope of the present article.