Hand Osteomyelitis: A Systematic Review of the Literature and Recommendations for Diagnosis and Management

Abstract

Hand osteomyelitis is a complex condition to diagnose and treat, with an opportunity to improve care through organization of existing evidence. The literature was systematically searched for series of hand osteomyelitis between 1990 and 2022 for evidence regarding diagnosis and treatment, to formulate recommendations. Twenty-one series reported at least 5 cases of hand osteomyelitis in adults, with a total of 666 cases. Surgical debridement is central to treatment and oral antibiotics are sufficient for individuals without diabetes, renal or vascular disease, after debridement and resolution of associated sepsis. A 4- to 6-week duration of antibiotic therapy according to organism sensitivities is recommended, or a 2-week course after amputation. Delayed presentation is common and if over 6 months is associated with high amputation rates. Hand osteomyelitis with renal failure is associated with systemic complications. Reconstruction options include antibiotic-eluting spacers, osteosynthesis or arthrodesis, vascularized bone or adipose, regional soft tissue coverage and silicone implant arthroplasty.

Level of Evidence: IV.

Introduction

Osteomyelitis in the small bones of the hand has specific challenges, with little evidence on diagnosis and treatment strategies. The structure of the nail complexes, tendon insertions, multiple small joints, and end-artery supply to the fingers provides a unique functionality, which differentiates it considerably from osteomyelitis of long bones and the axial skeleton, although some similarities with foot osteomyelitis exist. Presentation varies according to etiology, chronicity of disease, comorbidities, and route of infection, as does the resultant morbidity, surgical debridement, and reconstructive options. Early diagnosis and a combination of bone biopsy, surgical debridement, and pathogen-specific antimicrobial management have been proposed.^1,2 Suboptimal or delayed treatment may lead to long-term complications, or amputation.

The Cierny-Mader classification is widely accepted for osteomyelitis of the long bones. ³ In the hand, although many cases are identified on clinical and radiographic findings, an intraoperative assessment of whether the infection is medullary (stage 1), superficial (stage 2), localized (stage 3), or diffuse (stage 4) may aid treatment decisions. ² Most cases of hand osteomyelitis result from a contiguous focus of infection in the surrounding tissues. The joint, wound, tendon classification (JxWxTx) for hand septic arthritis includes an assessment of damage to osteochondral surfaces (no osteomyelitis is J0, periarticular osteomyelitis is J1), wound (W1 sinus, W2 purulent wound), or tendons (T1 extensor tendon destruction, T2 flexor, T3 both tendons) and highlights the reconstructive challenges. ⁴

A wide differential exists in terms of causative soft tissue bacterial infections in skin, wounds, and hand spaces. Mimics include rheumatological conditions and post-traumatic radiological appearances. Delayed presentation often contributes to the disease process. Outcomes range from complete recovery to progressive amputation, in an organ with crucial functional relevance which differs considerably between individuals according to employment, hobbies, and independence.

Recent developments in osteomyelitis research, including the oral versus intravenous antibiotics trial (OVIVA) have identified noninferiority of oral antibiotics after resolution of the acute episode, ⁵ with considerable health economic implications. ⁶ Septic arthritis in the hand, in the absence of osteomyelitis, usually requires a 2-week postsurgical course of targeted antibiotics, ⁷ as demonstrated by a prospective randomized noninferiority trial. ⁸ However, over 50% of the 99 hand and wrist cases studied had further mechanical or neurological sequelae. ⁸ The effectiveness of small joint washout and debridement, necessitating fewer washouts compared with larger joints, is suggested as a reason for the relative success of the shorter antibiotic duration. ⁹

Aims and Objectives

The aim of this systematic review was to evaluate the literature to identify and present the evidence base for treatment of osteomyelitis of the hand.

Methods

The results of this review are reported according to the PRISMA statement. ¹⁰

Eligibility Criteria

Eligible studies were case series and cohort comparisons published in the English language between 1990 and 2022. Eligible populations were adults with a clinical diagnosis of osteomyelitis of the phalanges or metacarpals of the hand. Series of metacarpal or phalangeal osteomyelitis as part of a larger series of hand infections, or as part of a series of osteomyelitis at any skeletal site, were also included. Septic arthritis cases were excluded unless osteomyelitis or radiographic changes within the bone was specifically described (Figure 1).

Figure 1.

Flow diagram demonstrating the search and selection process for the review.

Primary outcomes assessed were amputations and complication rates. Mean follow-up duration was elicited from each study. Individual study recommendations relating to the assessment and management of hand osteomyelitis were reviewed and described narratively.

Information Sources

The databases search included MEDLINE (Medical Literature Analysis and Retrieval System Online), CINAHL (Cumulative Index to Nursing and Allied Health Literature, Ipswich, Massachusetts), and EMBASE (Exerpta Medica Database, Amsterdam, The Netherlands). The search engine Google Scholar, and the Cochrane Database were searched separately over the same period. Reference lists were searched for further articles.

Search Strategy

MeSH terms and keywords included “hand,” and “osteomyelitis,” both of which were required for article retrieval. The search strategy was designed by the study’s principal investigator with the assistance of an institutional librarian.

Study Selection and Data Collection Process

Three reviewers (DD, MW, and MB) independently screened citations at title and abstract for eligibility. Articles considered eligible based on the title and abstract were retrieved and the full article screened for eligibility. Articles which could not be accessed were excluded. Data was extracted for demographics of participants, patient inclusion criteria, study design, and outcomes of interest.

Data Synthesis

Data was categorized and tabulated according to the series of hand osteomyelitis exclusively (Table 1) and series with hand osteomyelitis as a subgroup (Table 2).

Table 1.

Series With 5 or More Cases of Hand Osteomyelitis Published Between 1990 and 2022.

Series retrieved by literature review	Number of cases (n)	Cohort characteristics	Mean age (range)	Bones included	Amputations	Mean follow up	Other complications	Other outcome measures
Wyman et al, 2021, 12 Sheffield, UK	n = 210	Cohort A: hand OM and arterial calcification (n = 29)Cohort B: hand OM and DM and/or ESRF but no calcification (n = 43)Cohort C: all others (n = 138)	Cohort A: 62.5 (NR)Cohort B: 65.0 (NR)Cohort C: 54.6 (NR)	Metacarpals and phalanges	Cohort A: 25 of 29 (86%) phalanx amputations, 12 of 29 (41%) digit/ray amputationsCohort B: 17 of 43 (40%) phalanx amputations, 2 of 43 (5%) digit/ray amputationsCohort C: 40 of 138 (29%) phalanx amputations, 7 of 138 (5%) digit/ray amputations	NR		Cohort A: from the time of the diagnosis of osteomyelitis, all-cause mortality at 5 years was 69% (20 of 29)
Henry and Lundy, 2021, 15 TX, USA	n = 69	Direct inoculation hand OM	46 (10-74)	Phalanges and metacarpals	0 (2 excluded: uncontrolled DM; severe scleroderma)	16 weeks (±10)		All achieved resolution with oral antibiotics7 subsequent osteosynthesis procedures
Aimé et al, 2017, 21 Phoenix, AZ, USA	n = 12	Intra-articular OM treated with antibiotic-eluting, joint-spanning spacers	68 (37-87)	Phalanges and metacarpals	No amputations	22 months	Further debridement 1/12, bone loss around spacer 1/12, digital mal-alignment 1/12
Okada et al, 2015, 24 Osaka, Japan	n = 8	Chronic OM treated with pedicled adipose transfer	60 (29-80)	Phalanges	No amputations	41 months		No wound infection, dehiscence, sensory disturbance or recurrent OM
Gill and Lambah, 2014, 23 Dundee, Scotland, UK	n = 5	Delayed presentations >3 weeks	56 (52-63)	Phalanges	No amputations	NR	1 fracture nonunion1 PIPJ stiffness
Eisenschenk et al, 2005, 18 Berlin, Germany	n = 11	Carpals and MCs	43 (19-79)	Carpals and metacarpals	No amputations8 local or free flaps	Median 19.5 months (3 months -5 years)	1 death in hospital
Subasi et al, 2004, 25 Diyarbakir, Turkey	n = 7	Tuberculosis	22.7 (3-60)	Phalanges and metacarpals	No amputations	30 (16-52) months	1 MC shortening.
Reilly et al, 1997, 17 Cincinnati, OH, USA	n = 46 (6% of 700 hand infections)	Hand infections	46 (9-89)	Phalanges and metacarpals	18 of 46 (39%)If diagnosis >6 months, 86% amputation rate	43 weeks (2 weeks − 4 years)	Chronic pain 9 of 46, cold intolerance 2 of 46
Gonzalez et al, 1993, 19 Chicago, IL, USA	n = 24	Human bite OM	28 (20-42)	Phalanges and metacarpals (mostly at MCP)	1 of 24 (4%)2 of 24 K-wire joint stabilizations	At least 6 weeks		9 of 24 (38%) serial debridement
McLain et al, 1991, 20 Iowa City, IA, USA	n = 16 (of 146 hands)	Open fractures (all OM cases were Gustilo grade II or III)	33 (NR)	Phalanges and metacarpals	NR	NR	Function normal in 60% OM vs 88% of non-OMGrade III #s had poorer outcomes	Function: None—1Limited—4Functional—8Normal—3

Abbreviations: #, fracture; DM, diabetes mellitus; ESRF, end-stage renal failure; K-wire, Kirschner wire; MCs, metacarpals; MCP, metacarpophalangeal joints; NR, not reported; OM, osteomyelitis; PIPJ, proximal interphalangeal joint; RR, relative risk.

Table 2.

Series With a Subgroup of 5 or More Cases of Hand Osteomyelitis Published Between 1990 and 2022.

Series with subgroup of hand osteomyelitis 1990-2022	Number of cases (n)	Cohort characteristics	Mean age (range)	Bones included	Amputations	Other complications	Other outcome measures
Lipatov et al, 2022, 16 Moscow, Russia	n = 98 (of 170 patients)	Septic arthritis	NR for OM49 (IQR 34-65) for septic arthritis	Phalanges and metacarpals	2	7 required rehospitalization for sepsis	Median duration of antibiotics in osteomyelitis was 23 days, vs 7 days in septic arthritisROM (% of contralateral finger), MCP 64%, PIP 62%, DIP 59%
Gibson et al, 2024, 32 St Louis, MO, USA (accessed 2022)	n = 57 (of 145 patients)	Upper extremity infections in patients with type 1 and type 2 diabetes involving at least 1 digit	NR for OM	Phalanges	41 of 57 (67%)	NR	NR for OM
Butt et al, 2020, 33 Sialkot, Pakistan	n = 14 (13 of 250 diabetics; 1 of 220 nondiabetics)	Type 2 diabetics and nondiabetics with soft tissue upper limb infections	NR for OM	NR	NR	NR	NR
Smith et al, 2019, 29 Townsville, Queensland, Australia	n = 7 (321 tooth knuckle injuries)	Tooth knuckle injuries, incl. OM	NR for OM	Hand injuries	OR for amputation in OM was 35 (5 amputations in 321 injuries overall)	NR for OM	OM major risk factor for amputationDelayed presentation and DM increased risk of serial debridement
Sharma et al, 2018, 31 St Louis, MO, and Philadelphia, PA, USA	n = 22* (322 hand and forearm infections)	Hand infections in diabetic pts	NR for OM	Hand, wrist, and forearm	NR for OM	NR for OM	OM in DM vs non-DM, RR 2.7
Henry, 2015, 41 Houston, TX, USA	n = 9 (distal phalanx defects)	Vascularized MFC recon of fingertip	43 (22-58) years	Distal phalanx	NR	NR	Pulp pinch function returned in allMean union in 8.6 weeksNo reinfection, resorption or wound breakdown
Xu et al, 2010, 27 Singapore	n = 5 (11% of 47 ESRF hand infections)	ESRF	NR for OM59 (36-85) for ESRF hand infections	NR	NR for OM (17 of 47, 36%, for ESRF hand infections)	(NR for OM)7 MI, 5 sepsis, 3 DIC2 required inotropic support	Mortality: 11 of 47 (23%) at 1 year16 of 47 (34%) at 5 years
Benson et al, 2006, 28 Chicago, IL, USA	n = 7 (6% of 111 bite injuries)	Dog and cat bites	NR for OM	NR	NR	NR	Bony debridement, PICC line, IV antibiotics for 6 weeks, 6 office visits, 10 hand therapy visitsEst. cost of care $77,730
Houshian et al, 2006, 35 Odense, Denmark***	n = 23 (of 418 hand infections)	Hand infections, includes children	NR for OMMedian 40 (1-93)	Hand and wrist	NR for OM (14 amputations in 418 hand infections)	NR	NR for OM (11 developed stiffness 5 underwent arthrodesis)
Weinzweig and Gonzalez, 2002, 34 Chicago, IL, USA	n = 8* (1.8% of 448 upper extremity infections)	Surgical infections	NR for OM28.2 (11-74)	Hand and upper extremity	NR	NR	NR for OMDetailed microbiological analysis of hand infections
Francel et al, 1990, 30 Baltimore, MD, USA	n = 8 (41 hand infections)	Diabetics and diabetic renal transplant recipients	NR for OM	NR	NR for OM34 amputations among 8 OM cases in renal transplant recipients, 100% of transplant pts	14 of 41 had major deformities or hand dysfunction	26 of 41pts (63% amputations overall)

^***From reference list.

Abbreviations: DIC, disseminated intravascular coagulation; DIP, distal interphalangeal joint; DM, diabetes mellitus; ESRF, end-stage renal failure; IV, intravenous; MC, metacarpal; MCP, metacarpophalangeal joint; MFC, medical femoral condyle; MI, myocardial infarction; NR, not reported; OM, osteomyelitis; OR, odds ratio; PICC peripherally inserted central catheter; PIP, proximal interphalangeal joint; ROM, range of motion; RR, relative risk.

Confidence in Cumulative Evidence

The lack of prospective studies, absence of heterogeneity and randomization mean that the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) certainty ratings for each of the recommendations are low, except for the noninferiority of oral antibiotics and the use of (magnetic resonance imaging) (MRI) if diagnosis is in doubt.

Results

Twenty-one series were identified using the search strategy, with 666 cases of hand osteomyelitis. The largest cohort included 210 cases of hand osteomyelitis over 12-years (2008-2019) at a UK major trauma center. This cohort was examined for the predictive values of inflammatory markers for amputation, ¹¹ the role of digital artery calcification, ¹² secondary Raynaud phenomenon, ¹³ and the most common microorganisms. ¹⁴

Henry and Lundy ¹⁵ validated the cost-effectiveness of oral antibiotic therapy for 69 patients with hand osteomyelitis in the United States. Detailed descriptions of 98 cases with associated septic arthritis, ¹⁶ 46 patients with hand osteomyelitis, ¹⁷ 11 cases of metacarpal and carpal osteomyelitis, ¹⁸ 24 cases associated with human bites, ¹⁹ a series of 146 hands with open fractures of which 16 developed infection, ²⁰ 12 cases treated using joint-spanning antimicrobial-eluting spacers, ²¹ 7 managed with vancomycin-eluting spacers, ²² 5 delayed presentations, ²³ 8 treated with pedicled adipose tissue, ²⁴ and 7 caused by Mycobacterium tuberculosis ²⁵ (Table 1).

The articles with subgroups of hand osteomyelitis included skeletal osteomyelitis at any site (approximately 15 of 300 cases), ²⁶ upper limb infections in patients with end-stage renal failure (5 of 47), ²⁷ dog and cat bites to the hand (7 of 111), ²⁸ injuries related to human teeth (7 of 321), ²⁹ hand infections in diabetic and diabetic renal transplant recipients (8 of 41), ³⁰ hand and forearm infections in diabetic and nondiabetic patients (22 of 322), ³¹ patients with type 1 and 2 diabetes (57 of 145), ³² patients with type 2 diabetes, ³³ hand osteomyelitis in 8 of 448 upper extremity infections, ³⁴ and osteomyelitis in 23 of 418 hand infections ³⁵ with outcomes in Table 2.

A systematic review of 1669 open fractures of the hand in 1206 patients (12 studies) identified a total of 77 infections, of which only 11 (14%) were deep infections requiring debridement, the rest were regarded as superficial, managed with antibiotics alone, ³⁶ however this review includes all 16 infections reported by McLain et al ²⁰ as “superficial,” suggesting that well-defined criteria could benefit reporting of hand bone infections.

Evidence-based guidelines related to osteomyelitis were identified from the search. The Infectious Diseases Society of America have published guidelines on diabetic foot infections, including diagnosis and management of osteomyelitis confirming key outcome differences between soft tissue and osteomyelitis infections ³⁷ and which overlap with more than 20 similar national guidelines for diabetic foot infections, of which 13 provide diagnostic recommendations on foot osteomyelitis, ³⁸ however no dedicated national hand osteomyelitis guidelines were identified. The American College of Radiologists recommend MRI with contrast as the modality of choice for assessment of osteomyelitis, in addition to plain radiographs. ³⁹ A useful decision-making flow chart is available in the osteomyelitis guidelines from South Korea, ⁴⁰ with a subgroup of superficial osteomyelitis, treatable with superficial debridement, biopsy and culture, and 2 weeks of antibiotics if soft tissue cover is sufficient.

Discussion

The published series of hand osteomyelitis are heterogenous, retrospective cohorts, which outline a variety of nonstandardized outcomes, often as a subgroup of a wider study. Articles focus on specific etiologies (human teeth, dog and cat bites), or patient risk factors (diabetes mellitus or renal transplant recipients), while others focus specific sites (metacarpals), disease patterns with specific organisms (mycobacteria) or specific surgical procedures or adjuncts (antibiotic-eluting carriers).

Treatment options range from serial debridement and amputation in some cases, reconstruction after infection eradication with local or free flaps, ¹⁸ osteosynthesis or vascularized bone transfer,^15,22,41 pedicled adipose tissue transfer, ²⁴ antibiotic-eluting, and joint-spanning spacers.^21,22

Complications reported included amputation, chronic pain, cold intolerance, ¹⁷ stiffness, fracture nonunion, and systemic events such as myocardial infarction, sepsis, inotropic support requirement, and disseminated intravascular coagulopathy, related to severe infection and comorbidities. ²⁷ Hand-specific complications and long-term functional outcomes were infrequently reported.

Diagnostic Criteria

Diagnostic criteria for hand osteomyelitis remain controversial. No dedicated national guidelines were identified, contrasting with numerous national guidelines for foot osteomyelitis. An international consensus group have defined fracture-related infection criteria, including pathognomonic and suggestive criteria. ⁴² Pathognomonic criteria include clinical findings of a fistula, sinus, or wound breakdown (with communication to bone or implant) and/or purulent drainage from the wound or presence of pus during surgery, while laboratory related include 2 deep cultures culturing the same organism or histological identification of organisms.^39,42,43 Suggestive criteria include other clinical findings, imaging, inflammatory markers, and a single organism identified (Figure 2).

Figure 2.

International consensus definition of fracture-related infection (FRI) (reproduced with permission from Elsevier, from Metsemakers et al ⁴² ).

The 14 series reporting hand osteomyelitis exclusively (Table 1) each used different diagnostic criteria. Wyman et al ¹² used a combination of clinical, microbiological, and radiological findings, whereas Henry and Lundy ¹⁵ included only cases which were “surgically proven,” with gross purulence and bone loss at initial debridement and radiographic evidence of bone loss. Lipatov et al ¹⁶ use intraoperative revision, morphological evaluation (histopathology) and radiography, with the latter the least useful in early disease: only 60 of 98 (61%) cases had radiographic changes, with 41% in symptom duration <14 days and 83% for >14 days. Reilly et al ¹⁷ comment that history, physical exam, plain radiographs, and open biopsy and culture were most helpful in establishing the diagnosis. Gonzalez et al ¹⁹ define inclusion based on severe soft tissue infection with secondary bone involvement. Okada et al ²⁴ used a combination of plain radiographs and MRI to make a diagnosis. Gill et al ²³ use a combination of clinical, radiographic, and microbiological findings. Subasi et al ²⁵ base their study on the clinical picture and radiographic findings, confirmed by open biopsy. Eisenschenk et al ¹⁸ and Aimé et al ²¹ based the diagnosis on clinical findings and restricted their series to those undergoing surgical intervention.

Microbiological Analysis

Staphylococcus aureus is the most common organism isolated, although coagulase negative staphylococci feature frequently and many infections are polymicrobial, while patterns of antimicrobial resistance follow national trends.^14,17

Surgical intervention facilitates bone biopsies, to identify the causative organism(s) and guide further treatment. Ideally, 2 or more separate bone samples with separate instruments are recommended, as well as a deep tissue sample, for microbiological culture. A separate bone sample for histopathological analysis is advisable.

Histopathological Analysis

Histopathological analysis may confirm the diagnosis, the chronicity of the osteomyelitis, and exclude differentials. Occasionally it may reveal Mycobacterium tuberculosis in nonendemic areas, ²⁵ or neoplasia such as squamous cell carcinoma at a discharging sinus. ¹⁷ Acute versus chronic osteomyelitis can be differentiated histologically, including necrotic areas of bone or cartilage. ¹⁶

Inflammatory Markers

Regarding the role of serum inflammatory markers in the diagnosis of hand osteomyelitis, a retrospective review of 146 cases revealed that C-reactive protein (CRP) was the most sensitive marker, and an increase in CRP between diagnosis and follow up was associated with an increased risk of amputation. Conversely, if both CRP and white blood cell count were normal at diagnosis, the risk of subsequent amputation was low. ¹¹

Magnetic Resonance Imaging

The role of MRI as a diagnostic adjunct in cases where gout is a differential, or with co-existing rheumatological conditions as are observed in secondary Raynaud phenomenon, ¹³ is established, and can differentiate based on presence of bone edema where diagnostic uncertainty exists. It is widely used and forms an integral part of guidelines for foot osteomyelitis.^34,37,38

Recommendations: Diagnosis

Based on the current available evidence, a multifaceted approach to the diagnosis of hand osteomyelitis is recommended. Osteomyelitis should be suspected if there is a history of soft tissue trauma, open fracture, recent surgery, or adjacent infection (such as septic arthritis), and clinical examination reveals tenderness, erythema, swelling, fever, or a discharging sinus over a bone. Clinicians should be particularly suspicious of osteomyelitis in patients with diabetes and renal failure. Initial investigations may include blood tests (CRP, white cell count, blood cultures if sepsis suspected) and wound swabs or tissue cultures for microbiological culture to identify resistant organisms. Radiographs of the hand remain the first-line imaging, which may demonstrate signs suggestive of osteomyelitis, such as osteolysis or periosteal reaction. MRI is useful if there is diagnostic uncertainty, particularly if symptoms are less than 14 days in duration. Alternatively, a diagnosis of hand osteomyelitis can be made if there is evidence of gross purulence and bone loss at surgical exploration of the wound. Sendi et al ⁷ propose 2 aspects: (1) clinical or radiological evidence of osteomyelitis plus and (2) microbiological or histopathological criteria in bone samples.

Surgical Management

The various treatment options available for the management of hand osteomyelitis have been summarized by Aimé et al (Figure 3). Digit amputation remains the curative option in recurrent or severe cases, however it results in permanent disability. In circumstances where digit salvage is preferable to amputation, periarticular osteomyelitis has been successfully treated with antibiotic-eluting methyl methacrylate spacers as a single-stage procedure. ²¹ Antibiotic-eluting spacers are an alternative to amputation in the reconstruction of proximal and distal interphalangeal joints affected by osteomyelitis, including cases with patients with poorly controlled type 2 diabetes and end-stage renal failure.

Figure 3.

Treatment options for osteomyelitis in the bones of the hand (reproduced with permission from Elsevier, from Aimé et al ²¹ ).

Regeneration and Reconstruction

Distal phalanx bone destruction with preservation of the soft tissue envelope may leave a dysfunctional pulp. One reconstructive option is vascularized bone grafting to the distal phalanx from the medial femoral condyle, with restoration of pulp pinch in a series of 9 patients, and mean time to union of 8.6 weeks. ⁴¹ Alternatively, bone regeneration may occur after antibiotic-eluting methyl methacrylate spacer insertion and removal. ²¹

Vascularized tissue transfer as a treatment for metacarpal and phalangeal osteomyelitis has been reported, although mostly as individual case reports. Okada et al ²⁴ describe transferring pedicled vascularized adipose tissue into bony defects after debridement of chronic osteomyelitis in 8 patients, which is hypothesized to eliminate dead space, improve blood flow, and improve antibiotic penetrance into the foci of osteomyelitis (Figure 4).

Figure 4.

Surgical technique for pedicled adipose tissue transfer, showing retrograde-flow pedicle (a and b), and anterograde-flow pedicle (c and d), with pedicle (P) and digital nerve (N) (reproduced with permission from Elsevier, from Okada et al ²⁴ ).

Recommendations: Management

Initial management of hand osteomyelitis should be urgent debridement of infected and necrotic tissue and washout. Antibiotics should ideally be started only after bone biopsies have been taken unless the patient is clinically septic. Empirical prescription advice may be sought from clinical microbiologists initially and again once the causative organism(s) is identified from deep tissue or bone culture results.

Subsequent management depends on the severity of infection, affected anatomical site and tissues, comorbidities, and patient preferences. Options include continuing with oral antibiotics for 4-6 weeks, removal of any associated prosthesis, serial debridement for recurrent infection or amputation. Reconstructive options will depend on the location and condition of the affected tissues, comorbidities, patient expectations and preferences, and the likelihood of success. If osteomyelitis is recurrent or severe, digit amputation may be the only curative option and should be considered early if digital vessel calcification is evident on radiographs.

Limitations

The search strategy selects mainly English language articles and is subject to bias reflecting the opinions of the authors. The articles included favor more severe osteomyelitis identifiable on radiographs, or at operation.

Conclusions

Osteomyelitis can be classified and prognosticated based on anatomy and host status. Hand osteomyelitis is associated with outcomes according to specific host, inoculation type, and organism factors. Diagnostic and inclusion criteria vary between series, although fracture-related infections have a consensus definition. Complications and functional outcomes in the hand appear to be under-reported in most studies. Standardization and increased reporting of series of hand osteomyelitis is necessary to generate the high-quality evidence needed to develop guidelines on this complex area of hand surgery.