Proposal for a pain therapy effectiveness index

PROPOSAL FOR A PAIN THERAPY EFFECTIVENESS INDEX  

Davide Corvi  

Anesthesiologist  

Palliative Care Specialist  

Don Gnocchi Group, Monza  

Given the vast body of studies on complementary therapies in pain management, particularly acupuncture, and the uncertainty of many results despite encouraging data suggesting potential effectiveness, considering the widespread use of the practice and the subjective perception of effectiveness by practitioners and patients, I believe it is appropriate to propose new investigation methods to refine and guide randomized, double-blind controlled trials. There is often a sense that such studies fail to hit the mark, not only due to the complexity and uncertainty of the subject matter but also due to inherent methodological flaws that may not always adequately capture the uniqueness of the doctor-patient relationship or the non-reproducibility of patients themselves.  

Moreover, reviewing the literature, it is not uncommon to find that large-scale studies free of methodological flaws, which are not later contradicted or downscaled by other studies or meta-analyses, are rare. On the other hand, patients cannot always afford to wait for medical science to definitively establish, perhaps decades or centuries from now, which approaches are unequivocally evidence-based for managing their pain, especially in frequent clinical situations where traditional protocols fail.  

The practice of acupuncture (as well as other complementary practices like hypnosis) is highly operator-dependent, and many uncertainties may stem from the inherent variability of the procedure. Therefore, I propose a Therapeutic Effectiveness Index (TEI) for pain management, enabling a preliminary study of a therapy focused on individual clinical cases. While not replacing large-scale studies, it could provide useful insights on its own.  

The following formula provides a numerical representation of the "likelihood that a therapy was effective," using the term "likelihood" in a more logical than statistical sense. Simply put, this number gives mathematical form to the "therapist’s sense and belief that a therapy was effective," while inviting the therapist to reflect on the logical and psychological reasons behind such a (well-founded or unfounded) belief.  

Although this article focuses on acupuncture, the index can be applied to any symptomatic therapy. The index does not claim to provide certainty but allows for valid comparisons between therapies. For example, if a therapist notices a very high index in treating headaches with acupuncture while others using pharmacological therapies report lower indices, this would already be a useful datum for researchers, who should strive (through classic randomized studies, but not exclusively) to explain the phenomenon.  

If such results were frequent, they would represent signals that the scientific community could no longer ignore, recordable in terms of quantity and frequency even before a randomized study achieves statistical significance. It is a common clinical experience that some therapies appear rapidly and unequivocally effective, and the clinical reasoning behind these observations relies on the conscious or unconscious analysis of certain variables by the practitioner.  

In my view, these variables are as follows:  

- Duration of the symptom before therapy (DS), expressed in hours.  

- Magnitude of improvement (M): (NRSpre - NRSpost), where NRS refers to the Numeric Rating Scale used to assess symptom intensity.  

- Confounding factors: For example, the concurrent use of another painkiller or undeniable psychological factors. If confounding factors are clear and indisputable, the index is no longer reliable and should not be used.  

- Time to effectiveness (Te): The time from the start of therapy to the perception of its effectiveness, expressed in hours. Clearly, the time to effectiveness cannot be shorter than the symptom’s frequency: a pain occurring every 24 hours requires at least 24 hours to assess the therapy’s effect.  

- Instantaneity (I): If the symptom disappears instantly during or a few seconds after treatment, I=5; otherwise, I=1.  

- Duration of therapeutic effect (DE): From a minimum of one hour to a maximum of one month (to prevent the index from increasing indefinitely over years, making it less useful for comparisons).  

- Symptom-free interval (IS): The average symptom-free interval (the inverse of frequency). If the symptom is continuous, it is conventionally set to 1.  

The formula I devised is as follows:  

IET = (DS * M * DE * I) / (Te * IS * 10000)  

(The factor 10000 is used only to obtain more manageable numbers.)  

Continuous pains are conventionally treated as discontinuous pains with hourly frequency.  

The simplest way to understand the formula is to consider that resolving a headache lasting one year carries much more weight than one lasting one hour (hence DS is in the numerator). Naturally, M is also in the numerator, representing the subjective assessment of benefit. The duration of effectiveness contributes to the clinical utility of the therapy and is thus in the numerator. The time to effectiveness, for obvious reasons, is in the denominator: if a headache disappeared after a year of treatment, no one would credit the therapy’s effectiveness (or perhaps they might, cautiously, if the headache had lasted 40 years).  

Examples of Application:  

1. Daily headache for 1 month, NRS 5, with symptom-free intervals of about 20 hours, completely resolved during treatment (NRS 0), with the effect persisting for 4 days, no confounding factors:  

   IET = ((30*24) * 5 * (4*24) * 5) / (24 * 20 * 10000) = 0.36  

2. For pain that resolves instantly, the time to effectiveness is set to 1 (it cannot be less than the symptom’s frequency, and frequency cannot be less than hourly by convention, even for continuous pain).  

3. Weekly gastralgia for 1 month, NRS 5, completely resolved without confounding factors (gone for 14 days, starting the day after therapy, with symptom-free intervals of 6 days):  

   IET = (30*24 * 5 * 14*24 * 1) / (24 * 6*24 * 10000) = 0.0035  

   As seen, this case provides far less certainty about therapeutic effectiveness, as common sense would suggest.  

4. Continuous low back pain for 7 days, NRS 10, disappears instantly after an acupuncture session, concurrently with ibuprofen intake:  

   IET: Not calculable due to confounding factors. (However, it could be used to evaluate the combined therapy’s effectiveness [acupuncture + ibuprofen].)  

5. Continuous low back pain for 7 days, NRS 10, disappears instantly after an acupuncture session, lasting 4 hours, no confounding factors:  

   IET = (7*24 * 10 * 4 * 5) / (1 * 1 * 10000) = 3.36  

   Here, the time to effectiveness is conventionally set to 1, as is IS, as explained above. This case appears the most effective, as common sense might also suggest. A single number makes it easier to quantify the "belief that a therapy was effective."  

How to proceed in case analysis?  

Even if it were a mere placebo effect, such a result cannot be ignored, and it would be necessary to investigate why, among various placebo effects, this one was so effective as to surpass common analgesics (e.g., an NSAID would not have acted so quickly). It would then be necessary to test the same therapy by the same acupuncturist on multiple patients, compare the indices with other therapies, and only then, after clarifying protocols, proceed to randomized controlled trials. If such a study failed to confirm the therapy’s success, it would be worth studying the specific acupuncturist’s practice further, in case their unique skill was lost in larger studies with multiple operators.  

In short, case analysis is bidirectional: from the clinical case to the randomized study and back to the clinical case to refine larger studies.  

Potential applications of this index:  

1. It provides a numerical approach that replaces the practitioner’s subjectivity in assessing therapy effectiveness, focusing on the set of factors.  

2. It acts as a filter to evaluate which therapies are worth pursuing in future randomized studies, selecting the best therapies and increasing the likelihood of high-quality results.  

3. It identifies particularly effective practitioners: if one practitioner consistently reports higher IETs than others for the same therapy, their practice could be studied to identify factors contributing to the discrepancy, potentially uncovering effective elements not previously considered, thus standardizing procedures.  

4. It helps identify patient types most responsive to a therapy, allowing subsequent studies to focus on these subgroups.  

5. As shown by Kiene, randomized trials primarily investigate one type of cause-effect relationship in medicine. For some cause-effect studies, a single clinical case can be more significant.  

First Real Clinical Case  

45-year-old woman, continuous nausea and headache (duration 24 hours, headache NRS 8, nausea NRS 7). Auricular stimulation with Vaccaria seeds (on Shenmen, cranial tender points, and stomach tender point). Immediate headache relief (NRS 2 for over 4 hours), no nausea relief (NRS 6).  

IET (headache) = 24 * 6 * 4 * 5 / (1 * 1 * 10000) = 0.288  

Second Real Clinical Case  

NRS 8 dyspnea, NRS 8 continuous right-sided chest and back pain for at least 24 hours. Lung-specific points, dorsal Ashi points + 40BL bloodletting + 60BL right, 3K left. At session’s end: NRS 4 dyspnea, NRS 0 back, NRS 0 chest. Well-being (NRS 0) for 36 hours.  

IET (pain) = (24 * 8 * 36 * 5) / (1 * 1 * 10000) = 3.46  

IET (dyspnea) = (24 * 4 * 36 * 5) / (1 * 1 * 10000) = 1.7  

Third Real Clinical Case  

Headache pain for a few hours (duration not quantified, estimated at 3 hours) in a patient with brain metastases during a post-immunotherapy febrile episode. Morphine drops taken 15 minutes prior. Frontal headache radiating to the ear, NRS 6. BL3, VG20, BL2, LI4. Pain resolved after 20 minutes.  

To evaluate the combined morphine + acupuncture intervention: Unfortunately, the duration of the effect is missing and cannot be reconstructed due to the patient’s poor memory, so the IET cannot be calculated.  

PROBLEMATIC ISSUES  

In cases of instantaneous symptom resolution, a paradox may arise with two distinct effectiveness indices: for example, for a symptom lasting one month with daily frequency, the index could be calculated for the entire month or the single acupuncture session, yielding different numbers. However, this is not a significant issue: it suffices to clarify which choice was made when comparing with other therapies. These represent two distinct problems: chronic pain and its prevention versus acute pain and its resolution. Having two indices is not misleading.  

Another concern may arise for therapies requiring multiple sessions, which may yield a low index but still be effective in some way. What matters, however, is the comparison with other therapies for the same clinical issue: the index enables comparison, which is more important than the absolute value itself.  

DISCUSSION  

The Therapeutic Effectiveness Index offers, for the first time, a numerical quantification of a causal relationship under evaluation. How can we establish whether a therapy is the true cause of a recovery? More broadly, how can we determine if there is a causal link between two events? The philosophy behind cause-and-effect studies is beyond the scope of this article, but I recommend reading Kiene’s monograph (1). From a statistical and logical perspective, it is critical to distinguish simple correlation from true causality, although, in reality, causality is never linear, as multiple co-causes, not a single cause, produce an effect.  

For clinical purposes, it is sufficient to establish a certain correlation between a performed intervention and a result, and the parameters listed above are the most commonsense approach to this issue. Naturally, this is not the only way to combine these variables, and I hope the scientific community will refine this or similar indices to better suit research goals.  

It would be beneficial for therapists, especially those invested in promoting complementary medicine in the scientific community, to compile tables with effectiveness indices for various symptoms and patients treated, enabling comparison with colleagues and creating data collection platforms. Such comparisons would foster continuous improvement among therapists.  

Finally, it is worth considering the logical and statistical evaluation of causal links. It seems unlikely that a chronic pain would spontaneously resolve suddenly, especially if no changes in the patient’s life are detectable. In such cases, an external intervention temporally correlated with recovery strongly suggests causality. If this inferred causality repeatedly yields a high IET, it would be unscientific to ignore this data and rely solely on randomized trials and meta-analyses.  

Given a voluntary therapeutic act, what is the likelihood that recovery follows shortly after? The higher the index, the lower the probability that the recovery occurred by chance, unrelated to the therapeutic act. This cause-effect relationship must be considered, even if never fully provable and belonging to the realm of "weak causality." Repeated recovery phenomena make a high index even more suggestive: if a patient recovers every time they undergo acupuncture, not just in a single event, the causality becomes even more evident.  

Davide Corvi

davidecorvi@gmail.com

Reference:  

1)Kiene H, Hamre HJ, Kienle GS. In support of clinical case reports: a system of causality assessment. Glob Adv Health Med. 2013 Mar;2(2):64-75. doi: 10.7453/gahmj.2012.061. PMID: 24416665; PMCID: PMC3833527.