Research into the benefit that neurosurgical procedures bring to patients with degenerative spinal conditions has come a long way, from an initial focus on radiological outcomes (such as fusion rates) and physician-rated patient outcome (such as Nurick scale) to patient-reported outcomes—supplemented by tests of objective function designed to supplement the aforementioned subjective patient-reported outcomes with a more transparent, repeatable, physical performance component [1]. This multidimensional coverage of the various aspects of how a patient is doing after surgery already does a formidable job at capturing well-being and recovery trajectories. Can we perhaps do even better?

In the end, the rationale behind evidence-based medicine is to justify our medical and surgical decisions with the goal of maximizing their effect and maximizing the patient’s gain in well-being due to the intervention, while keeping a minimal treatment-related risk. Most of the evidence we act on in daily clinical practice is based on studies of efficacy (i.e., the performance of a certain intervention in optimal, at times artificial, study settings). These study settings, for example in a randomized clinical trial, ensure clear interpretability of results as well as comparability and generalizability, but only within the strict parameters set by the study (i.e., very strict inclusion criteria, strongly protocolized treatment paths, and pre-defined outcome measures as well as success definitions).

However, due to these strict parameters, it is often hard to apply the findings from such studies to any individual real-world patient in a real-world setting (effectiveness), as they cannot necessarily be extrapolated to the variety of clinical situations we meet every day [2]. After all, every patient with sciatica walking into your office is in some sense unique.

Pragmatic approaches to evidence synthesis have therefore gained traction in recent years, trying to test effectiveness (real-world performance) instead of efficacy (performance in a study environment). One aspect of pragmatic research is choosing outcomes and success definitions that are relevant to patients and sensitive to the “subjective” improvements (or deteriorations) that patients experience.

It would intuitively make sense that a more or less continuous measurement—that is taken throughout the patient’s day, rather than only at a specific timepoint/“snapshot” in a study setting (filling in a questionnaire, doing the planned six-minute walking test, etc.)—would be more reflective of “true” recovery (which we try to capture/model). This is what Ziga et al. [3] sought out to examine in their prospective study: Does adding a continuous measurement of physical activity give us a more sensitive model of recovery trajectories?

The authors’ group has already long been a trailblazer in the field of objective outcome measurements, both in the form of establishing and systematically analyzing objective functional tests such as the 6-minute walking test, Timed Up and Go test, and 5-repetition sit to stand test [4] as well as in accelerometer-based measurement of continuous physical activity using step counts [5]. In the present prospective multicenter study, Ziga et al. [3] followed 50 patients undergoing surgery for sciatica (due to lumbar disc herniation or recess stenosis due to bulging discs) at several timepoints up to 6 months postoperatively. Elegantly, normative data was used to derive age- and sex-adjusted normalized recovery trajectories, expressed in z-scores for each of the outcome measures, which included a battery of patient-reported outcome measures as well as the 6-minute walking test and smartphone-based step daily step counts. The authors not only visually but also statistically (comparing the recovery slopes over time using mixed models) show that the more traditionally “efficacy”-based measures (patient-reported questionnaires, objective functional tests) show a ceiling effect after about 6 weeks and, on average, show little improvement after that point in time. In contrast, the continuous daily step count, which is a more “effectiveness”-based measure, still showed a quite relevant improvement up to 6 months. This could either be explained mechanistically by these measured dimensions simply being very different, and some of these measures burning out after several weeks. On the other hand, the further improvement in step count up to 6 months clearly signifies an ongoing improvement in health state, a clear change in the patient’s life, even after the six week mark. This aspect of the recovery trajectory is clearly not modelled by patient-reported outcomes or by objective functional tests. Another contributing factor, as mentioned by the authors, could be the fact that objective functional tests are a sort of high-stress-test compared to the subconsciously taken step count over the whole day. This further strengthens the effectiveness versus efficacy argument.

Does targeting pragmatic endpoints really matter? Rarely do we see any huge revolutions in terms of achieved outcomes nowadays, research is for the most part incremental in its benefit to patients and often (in spine surgery) revolves around minimizing risks while maintaining good outcomes, rather than around a breakthrough in outcomes. To me, this is exactly why specifically choosing the most sensitive outcomes is so particularly relevant in our specialty: Choosing measurements that matter to patients, that are anchored in the real-world and measure effectiveness rather than efficacy, that are interpreted using thresholds known to be clinically important to patients, and that are generalizable and easily implemented (such as the step tracking proposed here)—this is the way to go to capture even incremental improvements that we can make through patient selection, patient preparation and recovery optimization, and technical refinements. The study by Ziga et al. [3] represents a good use-case: If longer-term recovery and change from baseline is assessed using only patient-reported questionnaires or a single objective test, relevant further improvements would be missed. Because it is our duty to constantly—if only incrementally—improve the care we provide to our patients, it is only natural to incrementally improve our clinical research methods. The study by Ziga et al. [3] represents such an incremental but important piece of the puzzle: Focus on real-world effectiveness and multidimensional assessment, and not only on efficacy.

Citations to this article as recorded by