Prosthetic implants

Medical implants have a detrimental effect on host defences such that a lower bacterial count is needed to initiate infection. Hence, there is a greater risk of infection during implant surgery. Bacteria growing on an abiotic surface, such as a prosthetic hip implant or heart valve, together with a protective layer of microbial polymers are known as a biofilm (Donlan and Costerton, 2002). Antimicrobials are frequently ineffective against micro-organisms growing in biofilms, making treatment of implant infections problematic and their prevention even more important.

Duration of surgery

The longer the operation, the greater is the risk of wound infection. This, in turn, may be influenced by the experience (Fig. 39.2) speed and skill of the surgeon and is additional to the classification of the operation by risk of infection, for example, clean, contaminated, dirty or infected.

Fig. 39.2 Surgical site infection rate by grade of operator for orthopaedic surgery 2008 (Health Protection Scotland, 2009).

Reproduced with kind permission from Health Protection Scotland.

Patient related factors

A number of patient related factors are known to influence the likelihood of developing a surgical site infection and include the following:

Table 39.3 American Society of Anesthesiology (ASA) classification of physical status (Mangram et al., 1999)

For each surgical procedure, a score of 0–3 is allocated to represent the number of risk factors present. Patients with a score of 0 are at the lowest risk of developing a surgical site infection, while those with a score of 3 have the greatest risk (Table 39.4). Use of this risk index allows comparison of similar patient groups in terms of surgical site infection risk over time. The risk index is a significantly better predictor of surgical-wound risk than the traditional wound classification system and performs well across a broad range of operative procedures.

Table 39.4 Risk index based on presence of co-morbidity and duration of operation (Culver et al., 1991)

Other factors

There are a number of other risk factors that may increase the risk of a surgical site infection (Table 39.5) for an individual patient but the impact has not been quantified to the extent of those risk factors discussed above.

Table 39.5 Patient and operative risk factors for surgical site infection

Age

Increasing age is associated with an increased risk of surgical site infection. However, there is debate whether age serves simply as a marker for underlying disease or whether the decline in immune function with age is the significant factor. A study of 72,000 patients in the USA, which adjusted for hospital type, procedure, duration, wound class and physical status of the patient, showed a 1.1% increase in surgical site infection per year of age from the age of 18 to 65 years, but a 1.2% decrease in individuals over 65 years (Kaye et al., 2005). In contrast, the findings of the English surgical site infection surveillance scheme (Fig. 39.3) indicated that the chance of getting a surgical site infection were 37% higher for a 65-year-old person compared to a 45-year-old person (HPA, 2008).

Fig. 39.3 Relative risk of surgical site infection by age group, 2003–2007, adjusted for surgical category (reference age group <45 years) (HPA, 2008).

Reproduced with kind permission from Health Protection Agency.

Antimicrobial prophylaxis

In the early 1960s, it was demonstrated, using a guinea-pig model, that surgical-wound infection could be reduced by administration of an antimicrobial just before an incision was made, but the beneficial effect disappeared if antimicrobial administration was delayed by 3–4 h after the incision (Burke, 1961). Since then, many clinical trials have indicated the benefit of maintaining adequate antimicrobial tissue levels from the time of initial surgical incision until closure.

There are potential adverse consequences to the administration of antimicrobials for both the individual and the population. For the individual, side effects, ranging from antimicrobial associated diarrhoea or thrush to life-threatening allergic reactions, may arise. From the population perspective, the development of antimicrobial resistant bacteria is a concern. Antimicrobial prophylaxis should, therefore, only be offered to patients where there is evidence or, in the absence of evidence, expert consensus that the potential benefits of prophylaxis outweigh the risks.

The number of patients that need to be treated with antimicrobial agents to prevent one infection in the different types of surgery are presented in Table 39.8.

The infection risk associated with a particular surgical procedure and evidence of efficacy should be used to determine whether antimicrobial prophylaxis is to be administered. Not all surgical procedures warrant antimicrobial prophylaxis.

Choice of antimicrobial

Once it has been determined that antimicrobial prophylaxis is required, the next step is to select an appropriate agent(s). The choice of antimicrobial should take into account the following:

The majority of clinical trials that have demonstrated the benefit of antimicrobial prophylaxis are outdated, and probably do not reflect current surgical practice. First and second generation cephalosporins (cefazolin and cefuroxime) have been the mainstay of agents studied (Bratzler and Houck, 2004). There are advantages and disadvantages with using cephalosporins. The advantages include a low anaphylaxis risk, but they have the disadvantage of excessive or inadequate spectrum of cover depending on the operation (Morgan, 2006) and a strong association with Clostridium difficile infection. Many antimicrobials used in prophylaxis have not been extensively studied in clinical trials, but are selected on a theoretical basis of their antimicrobial spectrum (see Tables 39.9–39.11).

Table 39.9 Antimicrobial susceptibility of common pathogens

(adapted from SIGN, 2008)

Table 39.10 Micro-organisms commonly isolated from surgical site infections and prophylactic antimicrobials used in common surgical procedures (Prtak and Ridgway, 2009)

Timing and duration

Timing of antimicrobial administration is one of the most important aspects of prophylaxis regimens. Animal studies and latterly clinical observational studies have shown that prophylaxis is most effective when given immediately before an operation (within 30 min of induction of anaesthesia), so that antimicrobial activity is present for the duration of the operation and for about 4 h afterwards. Antimicrobials given too early prior to surgery are associated with prophylaxis failure, presumably because serum and tissue levels are not sustained during the surgical procedure. Similarly, for each hour antimicrobial administration was delayed after the start of the operation there was an increased rate of wound infection. This suggests bacterial replication, once commenced, cannot be eliminated by antimicrobial regimens designed for prophylaxis. The microbiological basis for these observations is likely to be that bacterial reproduction at a logarithmic rate follows a lag phase of relatively little increase in bacterial population. The lag phase for wound infection bacteria lasts typically 3–4 h. If bacteria inoculated into a wound can be killed or inhibited by antimicrobials given early, the immune system can kill the remaining organisms. However, if antimicrobials are given only when the growth curve has entered the logarithmic phase, the chances of successful prophylaxis are reduced.

Formerly, protocols for prophylaxis extended for several post-operative days. Now, single dose schedules are increasingly common with greater emphasis on ensuring immediate preoperative administration. As surgery may be delayed at short notice, sometimes between the time the patient leaves the ward and arrives at the theatre, it is sensible for the administration of antimicrobials to be transferred from ward staff to the operating team when prophylaxis can be given around the time of induction of anaesthesia.

The optimum time to administer prophylactic antimicrobials before incision is probably 30 min, but national recommendations vary from less than 30 min (SIGN, 2008) to 60 min (NICE, 2008a) prior to incision. Figure 39.5 represents the two major studies undertaken to identify the optimum time for administration of prophylaxis. Both studies determined that post-incision administration of antimicrobials significantly increased the risk of surgical site infection.

Fig. 39.5 Timing of antimicrobial prophylaxis administration and infection rate. Study A: The timing of prophylactic administration of antibiotics and the risk of surgical-wound infection (Classen et al., 1992). Study B: Trial to reduce antimicrobial prophylaxis errors

(Steinberg et al., 2009). Taken from Mandell et al. (2010). Reproduced with kind permission from Elsevier.

Historically, the only occasion where antimicrobial administration has been delayed to after the incision is Caesarian section, where antimicrobials are given after cross clamping the umbilical cord to prevent drug delivery to the neonate. However, it is recognised that this does not provide the mother with adequate tissue levels at the time of incision and two studies have shown that antimicrobials can be given safely before incision without adversely affecting the neonate (Sullivan et al., 2007; Thigpen et al., 2005).

When a tourniquet is used during orthopaedic procedures to minimise bleeding, the antimicrobial should be infused before inflating the tourniquet. This ensures adequate tissue levels are achieved at the site of surgery (Bratzler and Houck, 2004).

Certain practical issues should be considered when selecting an antimicrobial, for example, the requirement for intravenous infusion or safe intravenous bolus administration. An antimicrobial, which requires to be administered over a long period, for example, vancomycin 1 g over nearly 2 h, is much less likely to be given completely compared to teicoplanin, which is administered as a bolus.

To improve the timing of antimicrobial prophylaxis administration, the World Health Organisation (WHO) have introduced a question in their surgical safety checklist. The question ‘Has antimicrobial prophylaxis been given within the last 60 minutes?’ is to be asked aloud before incision.

Repeat doses

Although single dose prophylaxis regimens are widely advocated (DH/HPA, 2008; NICE, 2008a; SIGN, 2008), many surgeons continue to use prolonged courses of ‘prophylaxis’ often for several days, without a clear evidence base. For some procedures, the optimum duration of prophylaxis is not known and 24–48 h prophylaxis is considered acceptable, for example, for open heart surgery (SIGN, 2008).

Where single dose prophylactic regimens have been adopted, the need for dosage adjustment in patients with reduced ability to excrete the drug (usually due to renal impairment) becomes unnecessary. This is because it is unlikely that single doses will have significant dose related adverse effects and idiosyncratic reactions are dose independent. Although the half-life of many drugs used is relatively short (1–2 h in normal volunteers), surgical patients often have slower clearance of antimicrobials from the blood. This concept will probably also hold true for prophylactic regimens lasting up to 48 h.

There are some situations in which it is necessary to prescribe additional doses of antibiotics to achieve the aim of adequate tissue levels at the time of wound closure. The additional doses may be needed when there is significant blood loss (>1500 mL) as plasma is effectively diluted by intra-operative transfusions and fluid replacement. Long operations may also need extra antimicrobial doses during the operation, but additional doses post-operatively do not provide an additional prophylactic benefit.

Antimicrobial administration by hospital theatre staff has practical implications for the route of administration. Ward-based administration of prophylaxis can be given orally if appropriate preparations exist, but this is impractical in sedated or unconscious patients. The oral route tends to suffer from variable absorption, especially in the presence of anaesthetic premedication, and this also makes it unsatisfactory. The intravenous route is the most reliable way of ensuring effective serum levels and is the only route supported by a substantial body of evidence.

A schematic model for the tissue concentration time profile of an antimicrobial agent used to prevent surgical site infection is presented in Fig. 39.6. After an initial dose of the antimicrobial agent, tissue concentrations reach their peak rapidly, with a subsequent decline over time. The goal of prophylaxis is for the antimicrobial tissue concentration to remain above the minimum inhibitory concentration (MIC) for the specific pathogens at the time of incision and throughout the procedure. The antimicrobial should be readministered during prolonged procedures to prevent a period where tissue concentrations are below the MIC (grey area). Failure to readminister antimicrobials appropriately may result in a period during which the wound is vulnerable. Recommendations for peri-operative re-dosing schedules are presented in Table 39.12. General guidance is to repeat doses of antimicrobials at intervals of 1–2 half-lives.

Fig. 39.6 Schematic presentation of tissue antibiotic concentration over time (Mandell et al., 2010).

Reproduced with kind permission from Elsevier.

β-Lactam allergy

Penicillin and cephalosporin antibiotics have been the cornerstone of antimicrobial prophylaxis to prevent surgical site infections. Patients reported to be allergic to β-lactam antibiotics or other antimicrobials need to be carefully assessed, as alternatives may not be optimal. Alternatives are often glycopeptides, for example, teicoplanin or vancomycin, which are more expensive, often need to be given by infusion (vancomycin) and can increase selection for resistant bacteria. The prevalence of penicillin allergy in the general population is unknown. The incidence of self reported penicillin allergy ranges from 1% to 10%, with the frequency of life-threatening anaphylaxis estimated at 0.01–0.05% (or 1–5 in 10,000). More than 80% of patients with a self reported allergy to penicillin have no evidence of IgE antibodies on skin testing. Important details of an allergic reaction include signs, symptoms, severity, history of prior reaction, time course of allergic event, temporal proximity to administered drug, route of administration, other medication being taken and adverse events to other medication (Park and Li, 2005). Reactions to penicillins and other β-lactams occur because of allergy to the parent compound or the metabolites. The cross sensitivity between penicillins and cephalosporins is unknown, but has been variably reported to be up to 10%. Early cephalosporin preparations were contaminated with penicillins probably leading to an over estimate of cross sensitivity (Saxon et al., 1987). As the generation of the cephalosporin increases, the likelihood of cross sensitivity decreases (Pichichero and Casey, 2007). Those with a penicillin allergy showed an increased risk of allergic reaction to a first generation cephalosporin. First generation cephalosporins share a similar side chain to penicillin and amoxicillin. However, cross sensitivity to second and third generation cephalosporins was lower. The different side chains appear to play a more dominant role than the β-lactam ring in allergy.

Recent prospective studies have shown that the cross-reactivity to carbapenems and monobactams is very small. It is around 1% for imipenem and meropenem, and no cross-reaction has been reported for aztreonam (Frumin and Gallagher, 2009).

The increased use of penicillins rather than first or second generation cephalosporins for surgical site infection prophylaxis is increasing the potential for adverse reactions. In addition, the current nomenclature for penicillin combinations, for example, co-amoxiclav, can often make it more difficult for staff to recognise penicillin containing antimicrobials. Current guidance on the use of β-lactams in patients with penicillin allergy is detailed in Table 39.13.

Table 39.13 Guidance on the use of β-lactam antibiotics in patients with penicillin allergy recommended by BNF (Joint Formulary Committee, 2010)