Periprosthetic Joint Infection After THA

Periprosthetic joint infection (PJI) is one of the most devastating complications of total hip arthroplasty. Despite advances in prevention it remains a leading cause of early revision, and its impact on patient outcomes is profound.

Incidence. Infection follows 1-2% of primary THAs and 3-5% of revisions. The incidence is increasing despite improved prevention, likely because the patients are higher-risk, and treating PJI costs 3-4 times as much as the primary arthroplasty.

Patient factors. These should trigger an optimisation protocol before surgery; the targets are set out under Prevention Strategies.

• Modifiable: obesity, diabetes, smoking, malnutrition, immunosuppression (steroids, biologics) and active infection elsewhere
• Non-modifiable: rheumatoid arthritis, prior surgery at the same site, male sex and advanced age

Surgical factors. The operating room environment and surgical technique have a major impact on infection risk. Risk rises with prolonged operative time (over 120 minutes), wound complications (haematoma, dehiscence), blood transfusion, revision surgery, bilateral simultaneous procedures and allogenic bone grafting. The protective measures are covered under Prevention Strategies.

Routes of infection. Direct inoculation, through intraoperative contamination or a postoperative wound complication, is the most common route for early PJI and accounts for the majority of early infections. Haematogenous spread is the route linked to late PJI, from dental procedures, urinary tract infections, skin and soft-tissue infections or gastrointestinal sources such as endoscopy and diverticulitis. Contiguous spread from adjacent osteomyelitis or a pre-existing septic hip is rare.

Biofilm. Biofilm formation is the key pathophysiology of PJI. Bacteria adhere to the implant surface and produce an extracellular matrix that protects them from the immune system and from antibiotics, raising the MIC 1000-fold. Antibiotics cannot penetrate established biofilm, which is why implant removal is usually necessary for chronic PJI.

The biofilm develops in stages:

1. Adhesion (hours). Bacteria adhere to the implant, mediated by surface proteins and a conditioning layer; this is the critical window for prevention.
2. Proliferation (days). Bacteria multiply, form microcolonies and begin producing extracellular matrix.
3. Maturation (1-3 weeks). A mature biofilm with complex architecture and a protective matrix drastically reduces antibiotic penetration.
4. Dispersion (ongoing). Bacteria shed from the biofilm can seed new locations, with chronic inflammation and bone resorption.

View larger

View larger

The organisms. Organism virulence affects both the presentation and the success of treatment.

• Staphylococcus aureus - 30% of PJI, the most virulent
• Coagulase-negative staphylococci - 30%, low virulence
• Streptococcus species - 10%
• Enterococcus - 5-10%
• Gram-negative rods - 10%
• Polymicrobial - 10-20%
• Culture-negative - 7-10%, with the worst prognosis

Tsukayama's temporal classification divides infection by when it declares itself, and it directly guides the choice of treatment.

Early infections (Type I), and acute haematogenous infections (Type III) with under 3 weeks of symptoms, may be suitable for DAIR. Delayed or chronic infection (Type II) almost always requires exchange, one-stage or two-stage. The three-week and three-month boundaries are historical heuristics rather than biological laws (see Management Algorithm), and Type IV is discussed under Unexpected Positive Intraoperative Cultures.

McPherson staging grades the host and the limb. Host grade affects antibiotic duration and how aggressive treatment should be.

Extremity grade affects the surgical approach and the need for plastic surgery involvement.

The presentation depends on when the infection declares itself.

Early, under 3 months. Fever, wound drainage and erythema: clear signs of infection, with systemic symptoms.

Subacute, 3-24 months. Pain, stiffness and mechanical symptoms. The presentation is subtle and may mimic aseptic loosening, so any painful THA, especially one that was never pain-free after surgery or became painful after a pain-free interval, should be worked up for infection.

Late haematogenous, over 24 months. Acute pain in a previously well-functioning THA, with an identifiable source of infection elsewhere.

History. Establish:

• The timing of symptoms relative to surgery
• Previous function of the THA: was it ever pain-free?
• Postoperative wound complications
• Recent infections or procedures (dental, UTI)
• Risk factors: diabetes, immunosuppression, prior surgery
• Prior antibiotic use, which affects culture yield

Examination. Look for wound erythema, warmth and swelling, a sinus tract or wound drainage, and previous surgical scars. Pain with movement, even passive, and a reduced range from contracture due to chronic inflammation are findings, and systemic signs (fever, and sepsis in severe cases) may be present.

The 2018 MSIS/ICM definition. One major criterion is enough for definite PJI:

1. Two positive cultures of the same organism from separate samples
2. A sinus tract communicating with the prosthesis

Without a major criterion, the minor criteria are weighted and scored, not counted; do not revert to "four minor criteria". A preoperative score of 6 or more indicates infection, 2-5 is inconclusive and needs the intraoperative criteria added, and 0-1 means not infected.

Serum markers. ESR and CRP are the first-line blood tests. They are screening tests and not diagnostic alone: they can be raised by aseptic inflammation, with a sensitivity of 80-90% and specificity of 70-80%. Raised markers prompt further workup, but normal markers do not rule out PJI, especially with low-virulence organisms; their trend is more useful for monitoring the response to treatment. A full blood count with differential can be sent, though leukocytosis is less sensitive. D-dimer was newly incorporated into the 2018 definition, and its limitations are discussed with the evidence below.

Aspiration. Joint aspiration is critical for diagnosis. It is image-guided, by fluoroscopy or ultrasound, with a lateral approach preferred, and a minimum of 3 samples go for culture. The fluid is sent for:

• Cell count and differential
• Culture: aerobic, anaerobic, and fungal if indicated
• Alpha-defensin, with the highest accuracy (sensitivity 97%, specificity 97%), though no test is universally best
• Leukocyte esterase, a rapid point-of-care test
• Synovial CRP

Intraoperative sampling. This provides the definitive diagnosis and guides antibiotic selection. Take a minimum of 5 separate tissue samples, not swabs, from deep tissues in different locations, for aerobic, anaerobic and, if needed, fungal culture; extend incubation to 14 days if a slow-growing organism is suspected.

Frozen section. Over 10 PMN per HPF (5 fields at 400x) indicates infection, with a sensitivity of 80-85% and specificity of 95%. It is useful for intraoperative decisions and can guide conversion to two-stage when infection is unexpected.

View larger

Sonication. Culturing the sonicated removed components disrupts the biofilm and improves yield, particularly in culture-negative cases, but it is not universally available.

Imaging. Radiographs are the baseline for every painful THA, looking for loosening, osteolysis and cement fracture; PJI can cause rapid bone loss. Imaging cannot reliably distinguish PJI from aseptic loosening, since both cause bone resorption, implant loosening and radiolucency, so it guides surgical planning while laboratory diagnosis is essential.

• CT assesses bone loss and guides revision planning
• MRI is usually not helpful because of artefact, and rarely used
• Tc-99m bone scan is sensitive but not specific
• Labelled WBC scan is more specific for infection
• FDG-PET remains a research tool, with improving availability

View larger

View larger

View larger

View larger

The sequence.

1. Clinical suspicion: serum ESR and CRP
2. If raised, or suspicion is high: hip aspiration, off antibiotics if possible
3. Aspirate for cell count, culture and alpha-defensin
4. Apply the MSIS criteria
5. If PJI is confirmed, plan treatment on timing and organism
6. Intraoperative cultures confirm the diagnosis and guide antibiotic therapy

Stabilise first. Recognise sepsis, obtain blood and deep cultures promptly, and begin empiric therapy after sampling when clinically safe. In sepsis or rapidly progressive infection, collect cultures promptly and start treatment: diagnostic yield does not outrank resuscitation.

Phenotype the infection. Confirm it against the definition score, then define the organism and its susceptibility, the symptom duration, any sinus or abscess, component fixation, bone loss, the soft tissues and the host's reserve. These choose the strategy; the aim is the one that achieves source control at an acceptable burden to the patient.

DAIR. Consider it when the components are well fixed, modular exchange and thorough debridement are feasible, and symptom duration, organism and host support retention. An unmanageable sinus or abscess weighs against it. Shorter symptoms favour success but there is no absolute cut-off, and patients should be counselled with centre- and organism-specific outcomes.

One-stage or two-stage exchange. Both can control chronic hip PJI. One-stage is a single definitive debridement and reimplantation, chosen when complete debridement and immediate reconstruction are feasible and the microbiological and local environment is acceptable. Two-stage separates them with an interval spacer, chosen when interval source control, local antibiotic delivery or staged reconstruction offers an advantage worth two operations. Choose between them on organism certainty, bone loss, soft tissue, host reserve, reconstructive complexity and local expertise.

Salvage. A retained spacer, chronic suppression, resection arthroplasty or, rarely, amputation is selected when definitive reconstruction is unsafe, unwanted or unlikely to succeed.

Antimicrobials. The route, the biofilm-active combination and the duration follow the organism, the operation, any retained material, source control, renal and hepatic function and infectious-diseases input.

Reimplantation. No single ESR or CRP, aspiration, antibiotic holiday or elapsed week proves eradication. The decision rests on the clinical course, the wound, serial evidence, operative findings and shared judgement.

The choice of agent follows the organism and its sensitivities.

Around a two-stage revision. After stage 1, give 4-6 weeks of IV antibiotics chosen on culture and sensitivities, then continue oral antibiotics until stage 2, or stop them 2-6 weeks beforehand for an antibiotic holiday. After reimplantation, give 2-4 weeks IV followed by oral suppression for 3-6 months at least; some advocate 12 months for S. aureus and MRSA. Infectious-diseases consultation is recommended.

Stewardship. This is a global principle:

• Reserve agents such as linezolid, daptomycin and newer cephalosporins for resistant Gram-positives (MRSA/VRE), according to local sensitivities
• Monitor drug levels for vancomycin and aminoglycosides
• Engage the antimicrobial stewardship and infectious-diseases teams in every case

Culture-negative PJI. It has worse outcomes. The causes are prior antibiotics, fastidious organisms and biofilm. Treat empirically with a broad-spectrum regimen (vancomycin plus pip-tazo or meropenem), and consider extended culture incubation, molecular diagnostics or fungal cultures.

When to suspect it. Fungal PJI, usually Candida and occasionally Aspergillus, is uncommon, around 1% of PJI. Consider it in an immunocompromised or debilitated host, after prolonged broad-spectrum antibiotics, in culture-negative infection that fails to settle, or as part of a polymicrobial infection. It is typically indolent, and standard bacterial cultures may miss it, so request fungal cultures with extended incubation and consider sonication and molecular tests.

Why it is hard. Candida forms a biofilm on the implant just as bacteria do, so eradication almost always needs implant removal and DAIR rarely works. Antifungal penetration is limited and treatment is prolonged.

Treatment. Two-stage revision is the mainstay, with an antifungal-loaded spacer: amphotericin B is the agent usually added to cement, as azoles and echinocandins elute poorly. Systemic antifungal therapy starts with an echinocandin or a lipid amphotericin formulation and steps down to fluconazole for susceptible Candida, continued as a prolonged course guided by infectious diseases. Outcomes are worse than for bacterial PJI, and resection arthroplasty (Girdlestone) or long-term suppression is a realistic endpoint in the frail or multiply-failed patient.

The single-culture problem. At a revision done for presumed aseptic failure, one unexpected positive culture is usually a contaminant and does not by itself indicate infection. Two or more cultures growing the same organism (Tsukayama Type IV) is the threshold for a true, previously unrecognised low-grade infection. This is why at least five separate deep tissue samples are taken, and why a single positive result is interpreted with caution.

What to do. A genuine Type IV infection is treated with an organism-directed antibiotic course, commonly around 6 weeks and guided by infectious diseases, with the newly implanted components retained; reported outcomes are relatively good because the biofilm burden is low. A single positive culture in a well-appearing patient is generally observed with vigilance rather than treated, unless it is a virulent organism such as S. aureus or other criteria are met.

Positive cultures at stage 2. Growth at reimplantation is more concerning, because it suggests persistent infection. A single contaminant may again be observed, but two or more concordant cultures mean the infection was not cleared: continue or restart targeted antibiotics, reconsider whether reimplantation was premature, and be prepared for repeat resection or long-term suppression.

Recurrent infection is the most devastating complication. Resistant organisms, an immunocompromised host and inadequate debridement all raise the risk. It is managed by a repeat two-stage, with salvage by resection arthroplasty considered.

View larger

Antibiotic toxicity. Each class carries its own risk:

• Vancomycin - nephrotoxicity; monitor trough levels and creatinine
• Aminoglycosides - oto- and nephrotoxicity; avoid in renal impairment
• Rifampicin - hepatotoxicity and drug interactions; monitor LFTs
• Fluoroquinolones - tendon rupture; avoid in the elderly if possible

After DAIR. Recovery follows a standard THA, with the antibiotic course layered on top.

After stage 1. Weight-bearing is protected, partial with a walker or crutches, though an articulating spacer allows mobilisation. IV antibiotics start in hospital and continue at home through a PICC line. Manage expectations: this is a phase of significant disability, and the spacer is intended to be temporary, retention as definitive treatment being reserved for selected patients (see Outcomes and Prognosis).

Between the stages. Continue antibiotics per protocol, monitor the inflammatory markers and aspirate the hip off antibiotics before stage 2. Optimise the medical comorbidities and nutrition (albumin, vitamin D).

After stage 2. Standard THA precautions, early mobilisation on day 1-2, and weight-bearing as tolerated if fixation is stable; antibiotics follow the protocol under Antibiotic Therapy.

Long-term surveillance. Check the inflammatory markers at 3, 6 and 12 months, then review clinically and radiographically every year. Any new pain or symptoms should prompt a search for recurrence, and patients should be educated about dental prophylaxis.

Favourable factors. Sensitive organisms, Streptococcus best of all; a healthy host without immunosuppression; early diagnosis and treatment; a good soft-tissue envelope; and adequate bone stock.

Unfavourable factors. MRSA, resistant Gram-negatives and fungi; an immunosuppressed host (steroids, biologics); multiple prior surgeries; extensive bone loss; polymicrobial infection; and culture-negative PJI.

The long view. PJI survivors have increased mortality compared with matched controls, and their quality of life is significantly affected. The economic burden, from lost work and caregiver needs, is substantial, and depression and anxiety are common.

When the spacer stays. In selected patients, spacer retention can be definitive treatment.

View larger

View larger

System-level prevention. Globally, prevention also works at the level of whole systems:

• National surgical site infection surveillance programmes (e.g. NHSN in the US, the UK SSI surveillance service, ECDC HAI-Net) benchmark and drive down rates
• Standardised perioperative antibiotic prophylaxis protocols and the WHO Surgical Safety Checklist
• Preoperative optimisation pathways (glycaemic control, MRSA decolonisation, nutrition, smoking cessation) in high-volume arthroplasty centres

Global epidemiology:

• Primary THA PJI incidence ~1-2%; revision THA 3-5%
• Infection is consistently among the top 3 indications for THA revision worldwide, alongside aseptic loosening and dislocation/instability
• Absolute case numbers are rising as arthroplasty volume and patient age/comorbidity increase, even where the percentage rate is stable

Registry evidence (cumulative revision for infection):

Side-by-side guideline comparison:

Resource-setting variation:

• High-resource: dedicated bone-infection MDTs (orthopaedics, ID, microbiology, plastics), sonication, molecular/next-generation sequencing, articulating spacers, and single-stage exchange programmes
• Limited-resource: reliance on serum markers and aspiration without alpha-defensin or sonication; two-stage and resection arthroplasty (Girdlestone) more frequently used where revision implants, prolonged IV antibiotics, or spacer fabrication are constrained

Related pages: Periprosthetic Joint Infection is the parent topic covering principles common to all joints, and TKA Complications: Infection is the knee counterpart - the diagnostic thresholds differ between hip and knee, which is why the 2018 definition's synovial cut-offs must be applied to the right joint; Biofilm Formation is the reason tissue culture misses a third of cases, why sonication works, why rifampin is added, and why a chronically infected implant must come out - every treatment decision on this page descends from it; Orthopaedic Antibiotic Therapy and Antibiotic Pharmacology and Resistance for the agents, durations and the rifampin combination rule the IDSA card anchors; Local Antibiotic Delivery for cement spacers, beads and the elution kinetics that determine whether an interval spacer is therapeutic or merely a placeholder; Perioperative Antibiotics and Surgical Site Infection Prevention for the prophylaxis that prevents most of this; Revision THA for the reconstruction at reimplantation, including the bone loss a resection arthroplasty leaves behind; THA Aseptic Loosening is the differential that must be excluded in every painful replacement, because an infected loose stem and an aseptically loose stem are radiographically identical; Synovial Fluid and Synovium for the aspirate that carries five of the eight scored minor criteria; and Osteomyelitis Pathophysiology for bone infection without an implant, and for the sequestrum-and-involucrum logic that explains why dead bone and dead space must be addressed as well as the organism.

The evidence base in PJI is evolving, and several management questions remain genuinely contested.

One stage or two. Two-stage is the traditional "gold standard", and the success rates quoted above favour it. But the INFORM RCT (BMJ 2022) and the Kunutsor meta-analysis show comparable re-infection with faster recovery, fewer intraoperative events and lower cost for single-stage in selected hips. The meta-analysis pooled cohorts in which the surgeon chose the operation, and the debate is shifting from "which is better" to "which patient for which strategy".

Antibiotic duration. The OVIVA trial (NEJM 2019) showed oral therapy was non-inferior to IV for bone and joint infection in the first 6 weeks, challenging routine prolonged IV courses. Total duration, 6 weeks against 12 weeks against lifelong suppression, remains unsettled and is being tested in ongoing trials.

DAIR criteria and timing. The "3-week biofilm window" is a useful heuristic, not a hard rule. Reported success varies widely with the organism, the host and the quality of debridement, and whether a single thorough DAIR or a planned repeat debridement is better is debated.

Diagnostic thresholds. Alpha-defensin, synovial CRP, leukocyte esterase, D-dimer and next-generation sequencing each add value but lack universal cut-offs. Culture-negative PJI and the significance of low-virulence organisms such as Cutibacterium remain difficult areas.

Spacer type and dosing. Articulating against static spacers, and high-dose antibiotic loading of cement, are supported mainly by cohort data; the systemic toxicity ceiling and the ideal elution profile are not standardised.

Suppression or eradication. In frail or multiply-revised patients, long-term suppressive antibiotics with implant retention may be preferable to repeated major surgery, but their durability and the risk of resistance are uncertain.