In a 2008 review of neurologic infection after neuraxial anesthesia, Reynolds illustrated the difficulty of identifying risk factors for infection in the obstetric population. In a summation of studies comprising over 1 million obstetric anesthetics, the incidence of meningitis after spinal anesthesia was
1:39,000, and that of epidural abscess after epidural anesthesia was 1:303,000. The total number of infections was small enough that it was difficult to identify causative factors in the OB population (5). In the surgical literature, however, there is a clear relationship between epidural abscess and patient age, immunocompromise, and prolonged epidural catheterization (6). The possibility of spontaneous epidural abscess in the absence of epidural anesthesia should be considered (7), as should coincidental community-acquired meningitis; the causative organism, however, may be very helpful in determining the time of acquisition of infection (Table 26-2).

Hand hygiene has been recognized for well over a century and a half as an integral part of infection control in the health care setting. The integrity of sterile gloves worn during neuraxial anesthesia can never be guaranteed, and appropriate hand hygiene will decrease the bacterial inoculum should a glove be punctured or torn during the procedure. Rings and wristwatches should be removed before hand washing, which is the most effective when an antimicrobial cleanser containing alcohol is used.

Since the most common causative agent for epidural abscess is Staphylococcus aureus (8), appropriate skin preparation is essential. Numerous studies have shown the superiority of chlorhexidine–alcohol to povidone–iodine in reducing growth of S. aureus, and in the prevention of central line associated infections (9,10,11). Iodophor–alcohol solutions are similarly superior to povidone–iodine alone in both immediate skin disinfection as well as epidural catheter colonization (12). Chlorhexidine is not inactivated in the presence of organic material such as blood, and its penetration of the stratum corneum of the skin gives it a prolonged duration of action, and effectively kills bacteria within hair follicles and sebaceous glands (13). While the package insert for the most commonly used chlorhexidine–alcohol preparation specifically states that it is not to be used prior to lumbar puncture (14), both the ASA in its practice advisory for the prevention of infectious complications in neuraxial anesthesia (15) and the American Society of Regional Anesthesia and Pain Medicine (ASRA) (16) have recommended the routine use of chlorhexidine–alcohol solutions for skin preparation prior to neuraxial anesthesia. The failure to demonstrate any increase in the incidence of neurologic deficits above baseline after the use of chlorhexidine for skin preparation in spinal anesthesia provides further evidence of its safety (17). And while chlorhexidine has been shown to be more cytotoxic than povidone–iodine at low concentrations in an in vitro model, there is no difference in toxicity at clinically relevant concentrations. Following the manufacturer’s instructions to allow the solution to dry for 2 to 3 minutes after application can further minimize any risk of toxicity of chlorhexidine (18).

In 2012, it is frankly astonishing that masks are not always worn during neuraxial anesthesia. It has been demonstrated that wearing a face mask results in a marked reduction in the bacterial contamination of a surface in close proximity to the upper airway (19). Both ASA and ASRA recommend the routine use of a face mask during neuraxial anesthesia, (15,16) and this is in fact a requirement of the United States Centers for Disease Control and Prevention (CDC) (20).

Further, there are numerous reports of cases of meningitis after spinal anesthesia in which a mask was not worn during the procedure (21,22,23). In a noteworthy report from the CDC, two separate outbreaks of meningitis in five obstetric patients receiving neuraxial anesthesia were described (24). Three women in New York State who received CSEA from the same anesthesiologist were found to have the identical strain of Streptococcus salivarius, a common nasopharyngeal organism. While the anesthesiologist reported wearing a mask during the procedures, the hospital permitted unmasked visitors to freely enter labor rooms during neuraxial procedures. In Ohio, two women received single-shot spinal anesthetics from the same anesthesiologist, who was determined to have routinely performed neuraxial procedures without a mask; one of these women died. The causative organism in both cases was a strain of S. salivarius that was genetically identical to an organism cultured from the anesthesiologist’s nasopharynx. The need for the operator to wear a mask during neuraxial procedures cannot be overemphasized. Requiring any other personnel in the labor room during the neuraxial procedure to wear a mask, including family members, should be considered as well.

While the ASA has recommended that a gown be worn during invasive procedures performed in immunosuppressed patients, there is no evidence that this is beneficial or necessary in routine neuraxial anesthesia. It should be noted, however, that while the ASA makes no recommendations regarding the use of sterile gowns, a substantial minority (33%) of the participating consultants agreed or strongly agreed that gowns should in fact be worn during neuraxial anesthesia (15), and this practice is fairly common in the United Kingdom.

The integrity of the epidural infusion system must be maintained at all times; this is of particular importance in patients receiving postoperative analgesia on the postpartum ward, where the level of surveillance for breaks in sterility may not be as great as in a labor and delivery unit. Disconnections and reconnections should be minimized, and catheter removal should be seriously considered in the setting of an unwitnessed disconnection.

In-line antibacterial filters are commonly used in the United Kingdom (25) and the ASA suggests consideration of their use in the setting of long-term epidural catheterization (15). While there is in vitro evidence that these filters are effective in eliminating passage of bacteria contained within even highly contaminated solutions (26), clinical studies suggest that the use of filters does not decrease the incidence of catheter tip colonization (27), and CNS infections have been reported even when filters were used (28,29,30).

Bacterial contamination can occur at the time that epidural local anesthetic infusions are prepared, and while local anesthetics are weakly bacteriostatic (31,32,33,34), microbial growth can occur during prolonged infusion. It should be noted that the newer single-enantiomer agents, ropivacaine and levobupivacaine, have lower antimicrobial activity than the older racemic mixtures (35,36,37).

In 2004, the United States Pharmacopeia formulated regulations, USP Chapter 797, regarding the compounding of pharmaceutical preparations, including local anesthetic infusions for epidural administration. These regulations, which represent a national standard enforceable by the U.S. Food and Drug Administration (FDA), state boards of medicine, and the Joint Commission, mandate that local anesthetic infusions intended to be infused over several days be prepared under a laminar flow workbench (38,39,40). Unfortunately, these guidelines do not specifically address infusions administered over a shorter time frame. Nevertheless, a close reading of the regulations suggest that multidrug infusions (e.g., local anesthetic and opioid +/- epinephrine) are considered medium-risk preparations, and as such should be prepared under a laminar flow hood in a pharmacy (Table 26-3).

The two major infectious complications of neuraxial anesthesia are meningitis and epidural abscess. While there may be some overlap in their clinical presentations, they differ significantly in their incidence, risk factors, microbial etiology, pathophysiology, and treatment. An excellent review of these complications is provided by Reynolds (5).