Table 1

Characteristics of included studies with 50 or more patients

Study

Country

Year of study

Patient treatments

Number of participants

Mean age (% men)

Exclusive surgical method

Treatment

Overall follow-up

Outcomes

Deaths and losses to follow-up

N (%) reinfection at 2 years

Details


One-stage

Buchholz et al., 1981 [19]

Germany

1968 to 1977

Hip arthroplasty for OA (95%), others (5%)

N = 640

58.8 years (39.7%)

"Patients with deep infection involving arthroplasties of the hip" (p. 344).

Antibiotic-loaded cement

52 months

Need for further exchange, reinfection

90 deaths

99 (15.5%)

Reoperation due to "bad" outcome plus other infection-related outcomes

Loty et al., 1992 [20]

France

1980 to 1988

THR

N = 90

65.7 years (not specified)

"We usually manage infected total hip replacements by a one stage revision" (p. 330).

Systemic antibiotics and antibiotic-loaded cement

47 months ± 29 months

Reinfection, other failure, Merle d'Aubigné-Postel score

Four deaths, seven lost to follow-up

8 (8.9%)

Infections occurred between 6 and 24 months

Miley et al., 1982 [21]

USA

1969 to 1979

Hip surgery for fracture dislocation (47%), OA (36%), others (17%)

N = 100 (101 hips)

Men 56.2 years, women 59 years (53%)

"The operating surgeon must be prepared to perform either of these operations (1-stage or Girdlestone), depending on the surgical findings and medical work-up" (p79).

Intensive multiple-drug antimicrobial programme

48.5 months, minimum 32 months

Grading system focusing on quality of life and hip function. Grade IV outcome: drainage, constant pain, further surgery suggested

11 deaths

Eight hips (7.9%) Grade IV outcome

No information on reinfection within 2 years

Raut et al., 1995 [22]

Wroblewski, 1986 [24]

UK

1979 to 1990

Cemented primary THR (63%), revision THR (37%)

N = 183

64.5 years (48%)

"All infected arthroplasties during this period were treated with 1-stage reimplantation, irrespective of the organism that was cultured, unless the patient had poor quality bone stock." (p. 202. Wroblewski, 1986 [24])

Antibiotic-loaded cement

93 months, range 24 to 164 months

Persistent infection, Merle d'Aubigné-Postel score (pain, function and movement), radiology, re-revision, complications.

14 deaths (0 in year 1)

6 (3.3%)

Reinfection in first 2 years

Schneider, 1989 [23]

Switzerland

1973 to 1988

Hip implants

N = 72 (excluding 13 treated with irrigation alone)

Not specified

"Between 1980 and 1988, out of 42 deep infections...." (p. 527)

Data also included from 1973 onwards

Joint irrigation preceding revision

Follow-up interval not specified.

Reinfection (bad outcome)

19 (26.4%), rate from 1980 (16.1%)

No information on reinfection within 2 years

Two-stage

Chen et al., 2009 [25]

Taiwan

1993 to 2005

THR (excluding patients with proximal femoral allograft reconstruction)

N = 57

51.5 years (72%)

"[A]ccording to the protocol for short-term parenteral antibiotics therapy at this institution...." (p. 189).

Interim antibiotic-impregnated cement beads

67.2 months

Reinfection, Harris hip scores, radiological examinations

Five deaths and five lost to follow-up

7 (12.3%)

Reinfection in seven patients after first stage described in text

Fitzgerald and Jones, 1985 [26]

USA

1969 to 1979

Hip implant

N = 131

61 years (50%)

"[D]elayed reconstruction in 131 patients who had an infection after a previous total hip arthroplasty...." (p.828. McDonald et al. 1989 [32]).

Intensive

Cemented reconstruction with no added antibiotic

49 months; range 2 to 9 years

Reinfection

All hip implants

11 (8.4%)

Reinfection up to 429 days

McDonald et al., 1989 [32]

1969 to 1985

Specifically THR for OA (69%), fracture (13%), other (18%)

N = 81 (including additional 13 patients)

60.0 years (53%)

5.5 years, range 2.0 to 13.6 years

Reinfection, complications

THR only

6 (7.4%) estimated from survival curve

Haddad et al., 2000 [27]

UK

1988 to 1992

THR for OA (72%), other (28%)

N = 50

60 years (46%)

"[C]onsecutive patients all of whom were referred with an infected total hip replacement and treated using a standardised protocol" (p. 689).

Antibiotic-loaded beads and cement ball

Uncemented

5.8 years, range 2 to 8.7 years

Reinfection, Harris hip score, radiological outcome, complications

Two deaths

4 (8%)

No information on reinfection within 2 years

Hsieh et al., 2009 [28]

Taiwan

2002 to 2005

Prosthetic hip

N = 99

61 years (61%)

"The use of an ALCS in SEA for PHI has been a routine practice in our institution" (p. 93).

Antibiotic-loaded cement spacer

43 months, range 24 to 60 months

Reinfection, Merle d'Aubigné-Postel score (only in comparisons of antibiotic strategies), radiographic results

Three deaths, five lost to follow-up

8 (8.1%)

Reinfection between stages

Romanò et al., 2010 [29]

Italy

2000 to 2007

Hip prostheses

N = 102

58 years (34%)

"102 consecutive patients underwent two-stage revision of septic hip replacement" (p. 26).

Long-stem or short-stem preformed antibiotic-loaded cement spacers

Cementless

48 months

Reinfection, Harris hip score (only in comparisons of antibiotic strategies)

Three deaths, nine lost to follow-up

5 (4.9%)

Reinfection within 3 years (including between stages)

Stockley et al., 2008 [30]

UK

1991 to 2004

THR for OA (60%), posttraumatic arthritis (18%), others (23%)

N = 114

64 years (55%)

"[C]onsecutive patients with microbiologically-proven deep chronic infection of the hip were managed by a two-stage exchange procedure" (p.145).

Antibiotic-loaded cement beads

74 months, range 2 to 175 months

Reinfection

9 (7.9%)

Reinfection within 1 year

Toulson et al., 2009 [31]

USA

1989 to 2003

THR

N = 132

54.7 years (59%) in patients "who completed the entire protocol" (p.1054)

"All 132 cases of infected THAs treated at our institution...." (p.1052).

Spacer containing antibiotic impregnated cement used in 67%

64.8 months, range 24 to 203 months; eight patients followed for only an average of 7.2 months

Reinfection, Harris hip score (mean only)

34 deaths (2 with no infection information), 8 lost to follow-up

11 (8.3%)

Three new infections within 24 months, four infections not eradicated, four patients who died had infections

Combination of methods

De Man et al., 2011 [33]

Switzerland

1985 to 2004

THR

N = 79 hips

70 years (57%) in patients followed up

One-stage (n = 24), two-stage (n = 55)

Some patients had spacers between stages

Mean 3.8 years (SD 2.2)

Reinfection, Harris hip score including limping and walking, radiographic outcome

Seven lost to follow-up

Overall (n = 2, 2.5%)

One-stage (n = 1, 4.5%)

Two-stage (n = 1, 2.0%)

Elson, 1993 [34]

UK

Not specified

THR

N = 296 (definite or possible infection)

Not specified

One-stage (n = 235), two-stage (n = 61)

Antibiotic-loaded cement pellets used in two-stage method

Not specified

Reinfection, mechanical survival, radiological outcome

Overall (n = 36, 12.2%)

One-stage (n = 33, 14.0%)

Two-stage (n = 3, 4.9%)

Time of definite or possible reinfection (mean 25 months, range 1 to 68 months)

Ketterl et al., 1988 [35]

Germany

1976 to 1986

THR

N = 207

69 years (42%)

One-stage (n = 21), two-stage (n = 161), no reimplantation (n = 25)

Gentamicin cement

32 months

Reinfection, function

Overall (n = 24, 11.6%)

One-stage (n = 7, 33.3%)

Two-stage (n = 17, 10.6%)

Time of reinfection unclear

Ladero Morales et al., 1999 [36]

Spain

1985 to 1995

THR

N = 62 (another 12 with nonsurgical treatment)

74 years (53%)

One-stage (n = 2), two-stage (n = 37), resection-only (n = 23)

Some treated after introduction of perioperative antibiotic protocol

4.8 years

Reinfection, Merle d'Aubigné-Postel score, clinical and functional outcome (pain, mobility, gait)

Overall (n = 3, 4.8%)

One-stage (n = 0, 0%)

Two-stage (n = 3, 8.1%)

Time of reinfection unclear

Lecuire et al., 1999 [37]

France

1982 to 1997

THR

N = 57

70.6 years

One-stage (n = 16), two-stage (n = 41)

Uncemented

6.6 years

Reinfection, Merle d'Aubigné-Postel scale, Harris hip score

Overall (n = 2, 3.5%)

One-stage (n = 1, 6.3%)

Two-stage (n = 1, 2.4%)

Time of reinfection unclear

Oussedik et al., 2010 [38]

UK

1999 to 2002

THR

N = 50

65 years (42%)

One-stage (n = 11), two-stage (n = 39)

Antibiotic-loaded spacer used in two-stage

Gentamicin-loaded cement used in one-stage

6.8 years, range 5.5 to 8.8 years

Reinfection, Harris hip score, VAS satisfaction

Overall (n = 2, 4.0%)

One-stage (n = 0, 0%)

Two-stage (n = 2, 5.1%)

No information on reinfection within 2 years

Sanzén et al., 1988 [39]

Carlsson et al., 1978 [42]

Sweden

1974 to 1981

THR for OA (74%), congenital dislocation (8%), fracture (8%), others (10%)

N = 108 (110 hips)

64 years (53%)

One-stage (n = 78 hips), 2-stage (n = 32 hips)

In 44% of two-stage procedures, gentamicin-loaded PMMA beads were used

Gentamicin-loaded cement

71 months, range 24 to 117 months

Reinfection, loosening, function

Eight deaths within 24 months

Overall (n = 22, 20.0%)

One-stage (n = 17, 21.8%)

Two-stage (n = 8, 25.0%)

At least 22 of 25 reinfections in first year

Vielpeau and Lortat-Jacob, 2002 [40]

France

Up to December 1998

THR

N = 458 (including acetabular or femoral revision only, excluding methods with retention of components)

No age or sex details

One-stage (n = 127), two-stage (n = 222), resection (n = 81)

Antibiotic cement (n = 249), no antibiotic cement (n = 100) in one or two stages

Median 3 years, 81.5% followed for minimum of 2 years

Reinfection, complications

Overall (n = 72, 15.7%)

One-stage (n = 15, 11.8%)

Two-stage (n = 33, 14.9%)


ALCS, antibiotic-loaded cement spacer; OA, osteoarthritis; PHI, prosthetic hip infection; PMMA, polymethylmethacrylate; SEA, staged exchange arthroplasty; THA, total hip arthroplasty; THR, total hip replacement; VAS, Visual Analogue Scale.

Beswick et al. BMC Medicine 2012 10:18   doi:10.1186/1741-7015-10-18

Open Data