Truace
Subscribe

Integrity of artificial intelligence and machine learning evidence in spine surgery: a critical appraisal of systematic reviews

Objective Artificial intelligence (AI) and machine learning (ML) are increasingly integrated into spine surgery across diagnostic, perioperative, and predictive applications, driven by their potential to enhance accuracy, efficiency, and patient safety. As this literature on AI and ML in spine surgery rapidly expands, concerns have emerged regarding selective reporting and overemphasis on algorithmic performance without adequate attention to transparency, bias, or clinical applicability. Such practices raise the…

By the Truvace desk1 min read1 sourcePeer-reviewedTRV-2026-1662· cite & verify
Integrity of artificial intelligence and machine learning evidence in spine surgery: a critical appraisal of systematic reviews
The correct principles of treatment for angular curvature of the spine by Lee, Benjamin, 1833-1913. Public domain

In brief

Researchers searched PubMed, Scopus, and Embase for systematic reviews and meta-analyses of AI and machine learning in spine surgery. They narrowed 273 records to 37 eligible reviews and scored each for 15 types of spin and for methodological quality using AMSTAR 2.

The finding matters because systematic reviews strongly shape clinical interpretation, and inflated claims about algorithm performance could mislead decisions about diagnostic, perioperative, and predictive tools. The abstract describes association patterns but does not show that the spin changed patient care, so the downstream clinical effects remain untested.

Main points

  1. 32 of 37 reviews (86.5%) had spin in their abstracts, and every one of the 15 spin types was observed.
  2. The most common spin types were type 5 (51.35%), type 3 (48.65%), and type 9 (48.65%).
  3. Lower journal impact factors and reduced AMSTAR 2 confidence ratings were significantly associated with spin.
  4. The authors call for greater transparency, adherence to reporting standards, and methodological rigor.

The problem

Spin appeared in 86.5% of abstracts of AI and machine learning spine surgery systematic reviews, overstating conclusions and casting doubt on how reliably clinicians can interpret the evidence.

The rundown

The review followed PRISMA guidelines. After 90 duplicates and 124 ineligible titles or abstracts were removed, a further 22 studies were excluded, leaving 37 reviews for assessment.

Study characteristics recorded included PRISMA adherence, publication year, and level of evidence. Associations with spin were tested using chi-square, Mann-Whitney U, and Kruskal-Wallis tests and Spearman's rank correlation.

The authors say the literature has expanded rapidly, with concern about selective reporting and overemphasis on algorithmic performance rather than transparency, bias, or clinical applicability. They conclude that addressing these biases is critical for reliable evidence synthesis and informed clinical decision-making.

Sources

  1. Peer-reviewedJournal of Neurosurgery: Spine2026-10-09

The debate