2.1 Study design and literature search

A systematic review and pooled analysis was conducted to investigate post-COVID AVN of the hip. Relevant studies were identified through a comprehensive search of PubMed, Embase and Scopus databases, covering the period up to June 2024. Search term using MeSH terms separated by Boolean operators as: (((COVID-19) OR (SARS-CoV-2)) AND ((steroid) OR (glucocorticoid) OR (prednisolone)) AND ((avascular necrosis) (osteonecrosis) OR (AVN))). Reference lists of retrieved articles were also screened to identify additional eligible studies. The review followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Fig. 1). The protocol was registered in the International Prospective Register of Systematic Reviews (PROSPERO) with ID: CRD420251170807.

Close

Fig. 1 PRISMA flow diagram showing the study selection process for the systematic review.

Export to PPT

2.2 Inclusion and exclusion criteria

Studies were included if they met the following criteria: (1) reported cases of AVN of the hip in patients with confirmed COVID-19 infection, (2) provided information on patient demographics with steroid dose and duration, and (3) were published in English. Case reports, case series, and observational studies were considered. Exclusion criteria included studies without individual patient data, studies reporting only non-hip AVN, or studies focused solely on pre-pandemic AVN unrelated to COVID-19.

2.3 Data extraction

Data were extracted independently by two reviewers (AS, AR) using a standardized template. Extracted variables included: author, year, study type, number of patients and hips, sex, age, cumulative steroid dose (mg prednisolone equivalent), duration of steroid use, time from COVID-19 infection or steroid therapy to AVN diagnosis, AVN stage (Ficat-Arlet (FA) or Steinberg), and treatment modality. Discrepancies were resolved through discussion with the senior author (VKJ).

2.4 Outcome measures

The primary outcomes were: latency from COVID-19 infection to AVN detection (in days) and cumulative steroid exposure (mg prednisolone equivalent). Secondary outcomes included AVN stage at diagnosis and type of treatment (conservative, joint-preserving, or THA).

2.5 Risk of bias (quality) assessment

Quality will be assessed using the Joanna Briggs Institute (JBI) checklist for case reports and case series.⁷ Each domain (selection, ascertainment, causality, reporting) will be rated as low, moderate, or high risk of bias.

2.6 Data synthesis and statistical analysis

Quantitative data on latency and cumulative steroid dose were pooled using a random-effects meta-analysis with the inverse variance method. Mean values and standard deviations (SD) were calculated for continuous variables, and heterogeneity was assessed using the I² statistic. A significant heterogeneity was defined as I² > 50 %. Descriptive statistics were used to summarize patient demographics, AVN staging, and treatment approaches. All analyses were performed using Review Manager (RevMan, version 5.4) and R software.

2.7 Ethical considerations

As this study was a systematic review and pooled-analysis of published data, ethical approval was not required.

3.1 Study characteristics and demographics

A total of 17 studies involving 209 patients (313 hips) with post-COVID AVN of the hip were included in this review (Table 1).^8–24 The study designs comprised case reports (n = 10) and case series (n = 7). The age of patients ranged from 14 to 96 years, with a mean of 43.7 ± 16.2 years. There was a male predominance (n = 98) compared to females (n = 57) across the reported studies.

3.2 Latency of development of AVN

The latency from COVID infection or steroid administration to AVN diagnosis varied from 13.3 ± 3 days to 309 ± 45 days. Pooled analysis of nine studies reporting this outcome involving 206 patients yielded a summarized raw mean latency of 126.48 days (95 % CI: 95.5–157.46).^{8–11,13–15,21,23} Significant heterogeneity was observed across studies (p < 0.01, I² = 99 %), indicating variability in latency among patients (Fig. 2).

Close

Fig. 2 Forest plot of mean latency (days) from COVID-19 infection to AVN onset across nine studies (pooled mean: 126.48 days, 95 % CI: 95.5–157.46; I² = 99 %).

Export to PPT

3.3 Dose of steroid

On Polled analysis, the cumulative steroid dose was reported in eight studies comprising 193 patients.^8–13,17,22 The pooled-analysis demonstrated a summarized mean cumulative steroid dose of 1198.44 mg (95 % CI: 860.99–1535.88) with significant heterogeneity (p < 0.01, I² = 96 %) (Fig. 3).

Close

Fig. 3 Forest plot of cumulative corticosteroid dose (mg prednisolone equivalent) in post-COVID AVN across eight studies (pooled mean: 1198.44 mg, 95 % CI: 860.99–1535.88; I² = 96 %).

Export to PPT

3.4 Management of AVN

Regarding AVN staging, the majority of cases were reported as FA stage II (n = 73) or stage III (n = 38). Several patients presented with stage IV disease (n = 19), and some hips were unclassified due to incomplete reporting.¹² Treatment modalities were variable: core decompression was performed in 36 hips, total hip replacement (THR) in 42 hips, and conservative therapy including bisphosphonates, calcium, NSAIDs, and physiotherapy in 83 hips. Combination therapies, such as core decompression with BMAC injection or bisphosphonate therapy, were also reported.^19,20 Hyperbaric oxygen therapy (HBOT) and physiotherapy were utilized in isolated cases.²³

Overall, the data indicate a wide variability in latency, steroid exposure, and AVN staging, reflecting heterogeneity in clinical presentation and management strategies in post-COVID AVN patients.

3.5 Quality appraisal

All included studies underwent critical appraisal using the Joanna Briggs Institute (JBI) checklists for case reports and case series. The overall methodological quality of the included literature was satisfactory. Most studies clearly described patient demographics, clinical presentation, diagnostic approaches, and interventions. However, documentation of adverse events and long-term follow-up data was limited in several reports. All studies were deemed of adequate quality for inclusion in this review, as they provided sufficient clinical and imaging evidence to establish a temporal and causal relationship between COVID-19 infection, corticosteroid exposure, and subsequent development of avascular necrosis of the femoral head (Tables 2 and 3).

4.1 Steroid exposure and risk of AVN

Our analysis demonstrates that cumulative steroid doses varied widely across studies, from 45 mg to 1856 mg prednisolone equivalent, with administration durations ranging from 14 to 179 days. Studies with higher cumulative doses, such as Kandari et al. (2022) with 1856 mg, reported advanced AVN stages (stage IV in 8 hips), suggesting a dose-dependent effect of steroids on AVN severity.¹⁷ Conversely, lower doses, as in Agarwala SR et al., 2021, of 758 mg were associated predominantly with stage II disease.⁸ These findings align with prior reports indicating a direct correlation between cumulative steroid dose and AVN risk.^25,26

Our pooled data indicate a substantial cumulative steroid exposure in post-COVID AVN cases, with a mean dose of 1198.44 mg prednisolone equivalent. This aligns with existing literature highlighting the association between high-dose corticosteroid use and increased risk of AVN. Mont et al. (2015) reported a 3.6 % increase in AVN incidence for every 10 mg/day increase in corticosteroid dose, with doses exceeding 20 mg/day further elevating risk.²⁷ Similarly, a meta-analysis by Zhao et al. (2017) found that higher cumulative steroid doses and longer treatment durations were associated with increased osteonecrosis risk in SARS patients.²⁶

Our findings also suggest a significant latency period between COVID infection and AVN onset, averaging approximately 126 days. This delayed manifestation may complicate early diagnosis and intervention. The prolonged latency observed in our study is consistent with reports in other contexts, such as systemic lupus erythematosus, where AVN developed months after corticosteroid therapy.²⁸

4.2 Latency from COVID to AVN and pooled-analysis findings

Our pooled-analysis revealed a mean latency of 126.48 days (95 % CI: 95.5–157.46) from COVID-19 infection to the development of symptomatic AVN of the hip, with significant heterogeneity (I² = 99 %). This finding underscores the delayed onset of AVN following COVID-19, which may complicate early diagnosis and intervention.

Our observed latency period is consistent with several studies reporting delayed onset of AVN post-COVID-19. For instance, a study by Annapareddy et al. (2025) found that the average duration between COVID-19 diagnosis and AVN symptom onset was approximately 10.8 months.²⁹ Similarly, a study by Migliorini et al. (2024) reported a mean time of 80 days from COVID-19 infection to symptomatic AVN.³⁰

However, other studies have reported shorter latency periods. Seong et al. (2025) observed that AVN symptoms developed earlier in patients treated with both dexamethasone and methylprednisolone, with an average onset at 7.5 months compared to 12 months in those treated with dexamethasone alone.³¹ This suggests that the type and duration of corticosteroid therapy may influence the timing of AVN onset.

4.3 Cumulative steroid dose and pooled-analysis findings

Our pooled-analysis revealed a mean cumulative steroid dose of 1198.44 mg prednisolone equivalent (95 % CI: 860.99–1535.88) among post-COVID AVN cases, with significant heterogeneity (I² = 96 %). This aligns with existing literature indicating a dose-dependent relationship between corticosteroid exposure and the development of AVN.

A meta-analysis by Mont et al. (2015) demonstrated that each 10 mg/day increase in oral corticosteroid dose was associated with a 3.6 % increase in the incidence of AVN. Furthermore, doses exceeding 20 mg/day significantly elevated the risk of AVN.³²

Similarly, Zhao et al. (2017) conducted a dose-response meta-analysis involving 10 trials and found that higher cumulative doses and longer treatment durations of steroids were more likely to lead to the development of osteonecrosis in SARS patients. The summary relative risk (RR) of osteonecrosis was 1.57 (95 % CI 1.30–1.89, p < 0.001) per 5.0 g increase in the cumulative dose of steroids.²⁶

In the context of COVID-19, a study by Muthu et al. (2023) found that an increase in the cumulative dose of corticosteroid therapy was associated with an increased risk of developing osteonecrosis of the femoral head (OFH). The summary odds ratio (OR) was 1.16 (95 % CI 1.09–1.23, p < 0.001) per 2.0 g increase in cumulative corticosteroid usage.³³

4.4 Pathophysiological considerations

AVN of the hip, characterized by the death of bone tissue due to compromised blood supply, has emerged as a significant complication in patients recovering from COVID-19. The pathophysiology of post-COVID AVN appears to be multifactorial, involving both direct effects of the viral infection and iatrogenic factors, particularly corticosteroid therapy. Understanding these mechanisms is essential for timely diagnosis, risk stratification, and appropriate therapeutic intervention.

COVID-19 infection is associated with a hypercoagulable state and endothelial dysfunction, leading to microvascular thrombosis. These vascular changes can impair perfusion of the femoral head, creating ischemic conditions that predispose to osteonecrosis. Several reports have highlighted the increased incidence of AVN in post-COVID patients, suggesting that viral-induced microvascular injury may act as an independent risk factor for AVN development.³⁴

Corticosteroid therapy, frequently employed in the management of moderate-to-severe COVID-19, is a well-established contributor to AVN. Steroids may induce fat cell hypertrophy and fat embolism, increasing intraosseous pressure and compromising blood flow to the femoral head.³⁵ Additionally, corticosteroids inhibit vascular endothelial growth factor (VEGF) expression, impairing angiogenesis and reducing bone perfusion.³⁶ Steroids also induce apoptosis in osteocytes and osteoblasts, disrupting normal bone remodeling and repair mechanisms.³⁷ The cumulative steroid dose correlates strongly with AVN risk and severity, emphasizing the importance of judicious steroid use in COVID-19 management.³⁸

The interplay between COVID-19-induced vascular injury and steroid-induced bone toxicity likely potentiates the development of AVN. Microvascular compromise from the infection, combined with steroid-related intraosseous fat embolism and impaired angiogenesis, accelerates ischemic necrosis of the femoral head. This dual mechanism may explain the relatively rapid onset of AVN symptoms observed in post-COVID patients receiving corticosteroids.

Recognition of these pathophysiological mechanisms has important clinical implications. Physicians should maintain a high index of suspicion for AVN in post-COVID patients, particularly those with a history of corticosteroid therapy. Early imaging, especially magnetic resonance imaging (MRI), can detect AVN at reversible stages, enabling interventions such as core decompression or pharmacological therapy before progression to femoral head collapse. Preventive strategies, including careful monitoring of steroid dose and duration, may further reduce AVN risk.

4.5 Clinical implications

The emergence of post-COVID AVN of the hip presents significant clinical challenges, underscoring the need for early detection, rational corticosteroid use, and stage-specific management. Our pooled analysis demonstrated a mean latency of approximately 126 days from COVID-19 to AVN onset and a cumulative steroid exposure of 1198 mg prednisolone equivalent; findings that reinforce the urgency of timely surveillance and targeted intervention.

Early detection remains paramount in preventing irreversible joint damage. MRI screening within three to six months following COVID-19 recovery, particularly among patients who received corticosteroids or report persistent hip pain, enables the identification of early-stage AVN that may respond favorably to conservative measures. Such strategiesincluding bisphosphonate therapy, protected weight-bearing, and core decompression; have demonstrated efficacy in halting disease progression when initiated promptly. These recommendations align with the observations of Maheshwari et al. (2023), who emphasized the critical value of early diagnosis in reducing the need for THA.³⁹

Corticosteroid exposure remains the most consistent and modifiable risk factor associated with post-COVID AVN. Steroids induce fat embolism, increase intraosseous pressure, and impair angiogenesis—mechanisms that compromise femoral head vascularity and precipitate necrosis. The current analysis corroborates previous findings by Agarwala et al. (2022), who documented AVN onset within 2–6 months of COVID-19, often after high-dose or prolonged corticosteroid therapy.⁹Seong et al. (2025) further identified cumulative steroid dose as an independent predictor of both disease severity and earlier onset, particularly when dexamethasone and methylprednisolone were co-administered.⁴⁰ These data support minimizing total corticosteroid exposure and adhering to the lowest effective dose for the shortest necessary duration.

Heterogeneity across patient responses, reflected by the high I² values (96–99 %), suggests variability in individual susceptibility, influenced by comorbidities such as diabetes, hyperlipidemia, and coagulation abnormalities. Therefore, clinicians should exercise heightened vigilance in patients receiving >500–1000 mg prednisolone equivalent, especially those with metabolic or vascular risk factors.

Treatment strategies should be guided by disease staging. Our review found that most patients presented in early stages (FA I–II) and responded well to conservative measures, while advanced cases (III–IV) required THA. These findings mirror those of Velchov et al. (2023), who reported improved outcomes with early joint-preserving interventions and noted the necessity of arthroplasty in structurally compromised hips.²² Although THA provides pain relief and functional recovery, it carries long-term complications such as implant wear, dislocation, and infection; factors reinforcing the importance of early intervention and hip preservation whenever feasible.

From an economic perspective, early detection and conservative management offer substantial cost advantages. Maheshwari et al. (2023) highlighted that hip-preserving procedures such as core decompression and osteochondral grafting delay or prevent THA, reducing both direct surgical and indirect rehabilitation costs.³⁹

Finally, while corticosteroids remain indispensable in managing severe COVID-19, clinicians must balance their life-saving benefits against the long-term musculoskeletal risks. Preventive strategies, including prophylactic bisphosphonate therapy in high-dose users, may mitigate AVN incidence, as suggested by Banjaree et al. (2021).⁴¹

4.6 Strengths and limitations

This study provides a comprehensive synthesis of post-COVID AVN of the hip, incorporating 17 studies with detailed analysis of patient demographics, steroid exposure, and latency from COVID-19 infection to AVN onset, AVN staging, and treatment approaches. The inclusion of meta-analytic pooled data for latency and cumulative steroid dose allows for a quantitative assessment of risk factors, offering a clearer understanding of temporal and pharmacologic relationships in this emerging cohort. Furthermore, the study highlights the impact of cumulative steroid exposure on AVN progression, providing clinically actionable insights that can guide early diagnosis and risk mitigation strategies. The systematic presentation of treatment outcomes, ranging from conservative management to THA, facilitates direct comparison with existing literature and supports evidence-based decision-making.

Despite these strengths, several limitations should be acknowledged. First, the majority of included studies were case reports or small case series, introducing inherent selection and reporting biases that may limit the generalizability of the findings. Second, there was substantial heterogeneity in steroid dosing regimens, COVID-19 severity, follow-up duration, and AVN diagnostic criteria across studies, which may affect the pooled estimates of latency and cumulative steroid dose. Third, detailed information on confounding factors such as comorbidities, concomitant medications, and pre-existing vascular risk factors was inconsistently reported, limiting the ability to perform adjusted analyses. Finally, long-term outcomes beyond the short-to mid-term follow-up were not consistently available, which prevented a comprehensive assessment of the durability of treatment strategies, particularly joint-preserving interventions.

4.7 Future directions

The emergence of post-COVID AVN of the hip highlights critical gaps in our understanding of its pathogenesis, risk factors, and optimal management strategies. Future research should focus on large, multicenter, prospective cohort studies to better characterise the incidence, natural history, and risk stratification of post-COVID AVN. Standardised reporting of steroid exposure, COVID-19 severity, comorbidities, and imaging findings is essential to enable more robust meta-analyses and facilitate the development of evidence-based clinical guidelines.

Advanced imaging modalities, such as dynamic contrast-enhanced MRI or perfusion studies, could be explored to identify early ischemic changes in the femoral head, potentially allowing for preemptive interventions. Randomised controlled trials comparing joint-preserving procedures, pharmacologic therapies (bisphosphonates, anticoagulants), and timing of THA are needed to establish standardised treatment algorithms tailored to disease stage.

Additionally, mechanistic studies investigating the interplay between COVID-19-induced endothelial dysfunction, hypercoagulability, and corticosteroid-induced osteocyte apoptosis may elucidate novel therapeutic targets for prevention and early intervention. Finally, health economics research, similar to the framework proposed by Maheshwari et al. (2023), should assess the cost-effectiveness of early interventions versus late-stage THA, particularly in resource-limited settings, to optimize both patient outcomes and healthcare resource utilisation.³⁹ Collectively, these future directions aim to improve early detection, refine treatment strategies, and mitigate the long-term morbidity and socioeconomic burden associated with post-COVID AVN.