Osteonecrosis after surgical treatment for DDH is one of those complications that we all fear, but which we often discuss with a mixture of "sensations" and data that are not always aligned. The paper by Guo and colleagues (Frontiers in Pediatrics, April 2026) is interesting precisely because it tries to bring order with a systematic approach: it collects the literature on risk factors for AVN after (surgical) reduction and synthesizes it in a meta-analysis, distinguishing between univariate and multivariate analyses [1].
What they did (and why it matters)
The authors searched for studies on PubMed, Embase, Cochrane Library, and Web of Science up to May 1, 2025, including works that analyzed factors associated with AVN after surgical treatment for DDH [1]. In the end, 16 retrospective studies were included, totaling 1,631 patients and 1,941 hips, with 468 hips complicated by AVN [1].
It is important to remember: the evidence is all observational and retrospective, so the risk of confounding remains high. However, the numbers and method are solid enough to shift the discussion to "what weighs most" instead of remaining vague.
The three factors that remain in the multivariable analysis
The most useful part, for me, is the multivariable model: not many factors remain there, but those that do are clinically credible and often replicated.
- Absence of the ossification nucleus In the multivariable analysis, the absence of the nucleus is associated with a higher risk of AVN (OR 2.60; 95% CI 1.73–3.91) [1]. This result reignites an old debate: does the nucleus truly "protect"? Previous literature was contradictory, and the review acknowledges this; however, here the signal clearly reappears in the multivariable synthesis.
- More severe dislocations (IHDI III–IV) The extent of dislocation, measured by IHDI, also remains independently associated with the risk of AVN (OR 2.43; 95% CI 1.46–4.03) [1]. This is intuitive: greater dislocation often means more complex maneuvers, more "stretched" tissues, greater vascular vulnerability, and a higher probability of additional procedures.
- Secondary procedures Secondary procedures are associated with a higher risk (OR 2.56; 95% CI 1.02–6.46) [1]. Here, the interpretation must be done with caution: a "secondary procedure" can be a marker of a more difficult hip, a longer course, or a less stable/less congruent reduction. It is not necessarily the procedure itself that causes AVN; it could be the case profile.
What, on the other hand, is NOT so "solid"
It is interesting that various factors often cited in practice do not emerge as independently convincing in the multivariable synthesis (or at least not with a stable signal): for example, age (in some models), surgical approach, abduction angle (with high heterogeneity) [1].
This does not mean that they "never matter." It means that, based on the available data, they cannot establish themselves as primary drivers when all the literature is put together.
Practical Takeaway
This paper alone does not change DDH surgery, but it helps to organize counseling and follow-up:
- in cases with severe dislocation (IHDI III–IV) and/or absence of the nucleus, the risk of AVN is higher: this should be clearly stated and careful follow-up should be planned [1];
- the need for secondary procedures is a warning sign: it often indicates a more complex course and a greater cumulative risk [1];
- many "ancillary" variables (angles, technical details, immobilization choices) remain important, but to date, they are more difficult to isolate as independent factors with the available evidence.
Limitations worth remembering
- all included studies are retrospective [1];
- diagnostic criteria and classifications of AVN are not always uniform;
- confounding by severity (worse hips often receive more complex treatments) is inevitable;
- the OR for secondary procedures has a wide CI, indicating that the data is less precise [1].
In summary: a good meta-analysis does not serve to "choose the perfect technique," but to understand where the main risk lies. Here, the risk seems to be mainly biological and related to the severity of the dislocation, rather than a single isolated maneuver.
Disclaimer
This content is for informational purposes only and does not replace individual clinical evaluation.
References
[1] Guo Z, Chen X, Dai G, Xue M. Risk factors for avascular necrosis of the femoral head after surgical treatment of developmental dysplasia of the hip in children: a systematic review and meta-analysis. Frontiers in Pediatrics. 2026;14:1811138. doi:10.3389/fped.2026.1811138.
Link: https://www.frontiersin.org/journals/pediatrics/articles/10.3389/fped.2026.1811138/fulL
