Bacteremia associated with pressure ulcers: a prospective cohort study

Pressure ulcers are a frequent complication in patients with limited mobility. They mainly develop in elderly patients with debilitating diseases and in patients with spinal cord injuries. A prevalence of 11–25% in nursing homes and 3–11% in hospitals is estimated [1‐4]. The greater longevity of the population and the increase in spinal cord injuries due to traffic accidents has contributed to the increased incidence of pressure ulcers.

The main infectious complications that can develop from pressure ulcers are cellulitis, abscess, osteomyelitis, and bacteremia [5, 6]. Bacteremia is an infrequent complication but can lead to significant mortality [6, 7]. Published studies of bacteremia associated with pressure ulcers (BAPU) are scarce, and ulcers as the origin of the bacteremia are not always well-documented [8‐10]. On the other hand, large published case series of bloodstream infections do not usually specifically identify those originating from pressure ulcers, rather these cases are usually included in the group of bacteremia originating from skin and soft tissue [11‐14].

As part of our prospective study of all bloodstream infections, we have specifically studied the episodes of BAPU during a period of 32 years.

The bacteremia was hospital-acquired in 20 cases (35.7%), healthcare-related in 15 cases (26.8%), and community-acquired in 21 cases (37.5%). The baseline and clinical characteristics are summarized in Table 1.

Microorganisms isolated from blood culture and ulcer culture are detailed in Table 2. Bacteremia was polymicrobial in 14 cases (25.0%) while the ulcer culture was polymicrobial in 41 cases (73.2%). The most frequent causative agent of BAPU was Staphylococcus aureus followed by Proteus spp., Bacteroides spp., and Escherichia coli. Only 4 episodes of BAPU were caused by methicillin-resistant S. aureus (MRSA), all from 2002, and only one episode was caused by ESBL-producing Enterobacteriaceae. No bacterial isolates were carbapenem-resistant. Of the four episodes caused by Pseudomonas spp., two were hospital-acquired and two were community-acquired. The most frequently isolated microorganisms from the ulcer cultures were E. coli and Proteus spp., followed by S. aureus, Bacteroides spp., Pseudomonas spp., and Enterococcus spp.

In 49 cases (87.5%), the ulcers responsible for bacteremia were already present at admission, whereas in 7 cases (12.5%) they were acquired during the hospital stay. The percentage of hospital-acquired ulcers during the period from 1984 to 1999 (17.6%) was higher than in the period from 2000 to 2015 (4.5%), although the difference was not statistically significant (p = 0.22). In 42 cases (75.0%) there was more than one ulcer. The characteristics of the ulcers are shown in Table 3.

Empirical antibiotic treatment was appropriate in 44 cases (78.6%). All patients were treated locally with ulcer cleaning. Surgical debridement was performed in 36 cases (64.3%). Overall mortality was observed in 23 cases (41.1%). Death was considered directly related to bacteremia in 12 cases (21.4%). The age of the patients who died (77.5 ± 14.5 years) was not significantly different from the age of the patients who survived (74.7 ± 13.8 years; p = 0.46). Patients who died had significantly lower serum albumin (21.8 ± 4.9 g/L) than patients who survived (25.3 ± 4.0 g/L; p = 0.0067). Table 4 summarizes the risk factors for overall mortality rate. Variables associated with overall mortality in the univariate analysis were hospital-acquired bacteremia, polymicrobial bacteremia, failure to perform surgical debridement of the ulcer, and serum albumin below 23 g/L. Independent risk factors for mortality were hospital-acquired bacteremia, polymicrobial bacteremia, and serum albumin below 23 g/L.

When a patient with decubitus ulcers presents with fever in the absence of other foci of infection, bacteremia due to the ulcers should be suspected. On the other hand, the appearance of fever immediately after debridement of an ulcer may reflect the existence of transient bacteremia related to the debridement [10]. Although bacteremia is a well-known complication of pressure ulcers, its incidence is not well established and studies on this subject are scarce [8‐10]. The present study constitutes the largest case series of BAPU published so far. It has been prospectively obtained in a single institution, over 32 years, applying strict diagnostic criteria. In the whole study period, 1.7 episodes of BAPU per 10,000 hospital discharges were observed, which coincides with the incidence observed by Bryan [9]. It is noteworthy that in the period from 2000 to 2015, the incidence was significantly lower than in the period from 1984 to 1999 (1.05 vs. 2.78 episodes per 10,000 discharges). This reduction coincided with the implementation of a multidisciplinary team aimed at preventing and treating chronic ulcers, both in the hospital and in primary care centers dependent on our institution. Although only 12.5% of the ulcers were acquired in the hospital, 35.7% of the episodes of bacteremia were hospital-acquired. Hospital admission for an intercurrent illness is likely to be a risk factor for developing bacteremia from pre-existing ulcers, either due to further deterioration of the host defense mechanisms or changes in the microorganisms that colonize the ulcer.

A striking finding in this study was the high frequency of polymicrobial bacteremia, already observed in previous studies [8, 9, 19], as a consequence of the typical polymicrobial colonization of ulcers. S. aureus, which was the most frequently isolated microorganism, has a special ability to spread to the bloodstream from pressure ulcers. Thus, in 17 of the 20 cases in which S. aureus was recovered from the ulcers, this was the same microorganism that was responsible for the bacteremia. The same is not true of Enterococcus spp., which was recovered in 11 cases in ulcer culture but in only 2 cases was it responsible for the bacteremia. Among Enterobacteriaceae, Proteus spp. originating from pressure ulcers have a special ability to cause bacteremia and were responsible for more than a quarter of the episodes in this study. The prevalence of Proteus spp. has also been observed in other studies [8, 9], although a satisfactory explanation for this predominance has not yet been formulated. By contrast, while E. coli was recovered in the culture of ulcers in almost half of the cases, only in 8 cases was E. coli responsible for the bacteremia. Several researchers have highlighted the importance of anaerobic microorganisms, finding that the devitalized ulcer tissue provides an adequate environment for persistence and proliferation of such organisms [9, 20, 21]. It is worth noting the presence of Bacteroides fragilis, which reflects fecal contamination of ulcers, especially those in the sacral location [22]. This pathogen was responsible for 23% of the episodes of BAPU in this study.

We did not identify any clinical or epidemiological factors that allowed reliable prediction of blood culture findings. In addition, the history of a previous episode of BAPU does not appear to be useful in predicting the causative pathogen, as demonstrated in three patients in our study who presented for recurrent bacteremia due to a pathogen other than the one that caused the initial episode. This is because the local infection of the ulcers is usually polymicrobial and, likewise, the chances of colonization by new microorganisms are high. Therefore, when an episode of BAPU is suspected, empiric antibiotic treatment should adequately cover S. aureus, gram-negative enteric bacilli, and anaerobic microorganisms, including B. fragilis, taking into account the local resistance rates. Once the blood culture results become available, antibiotic treatment should be appropriately adjusted and directed to the microorganisms responsible for the bacteremia.

Pressure ulcers may constitute an important reservoir of multiresistant microorganisms, especially MRSA, with the consequent risk of bacteremia by this pathogen [23‐26]. It has been shown that the presence of a chronic wound is an independent risk factor for the persistence of MRSA colonization after hospital discharge [27]. In this study, only four episodes of BAPU were caused by MRSA. This may be due in part to the active search-and-destroy policy against MRSA adopted at our hospital where annual rates of methicillin resistance in S. aureus isolated from clinical samples in the last 10 years have ranged from 4.5 to 13.8% (unpublished data), lower than the percentages observed in most hospitals in our country [28].

Osteomyelitis is a well-known complication of pressure ulcers. It should be suspected in ulcers with a torpid course that do not improve despite intensive local treatment [29, 30]. A positive culture of a bone biopsy obtained through the ulcer may reflect only the colonization of the adjacent soft tissue, and the gold standard for the diagnosis of osteomyelitis is pathological examination of bone tissue [31]. Among non-invasive diagnostic methods, magnetic resonance imaging (MRI) offers the highest sensitivity and specificity [32]. A limitation of this study is that the presence of osteomyelitis was not systematically investigated, leaving the practice of MRI and bone biopsy to the discretion of the primary clinician. Therefore, it is possible that this complication has been underdiagnosed.

Overall mortality was high, although slightly lower than previously reported [8, 9]. In only half of the cases, death was directly related to bacteremia, a percentage similar to that previously described [9]. Hospital-acquired bacteremia, polymicrobial bacteremia, and serum albumin of less than 23 g/L were independent risk factors for 30-day mortality rate. In addition to being a marker of protein malnutrition, hypoalbuminemia may appear in a wide constellation of clinical situations, as a consequence of reduced hepatic synthesis, renal or intestinal losses, or transcapillary escape to the interstitial space [33, 34]. Albumin has been considered a negative acute-phase protein, since its serum concentration is prone to decrease in trauma and sepsis, and hypoalbuminaemia has proven to be a marker of poor prognosis in different clinical settings, including bacteremia [35, 36].

Surgical debridement of the ulcer is an important element in the management of BAPU. It is a safe procedure in patients with advanced ulcers, despite the comorbidities that these patients usually present [37]. Galpin et al. [8] observed cases of persistent bacteremia that were only resolved after surgical ulcer debridement, as well as a lower mortality among patients undergoing surgical debridement, as occurred in this study.

Finally, although broad-spectrum antibiotic treatment and surgical debridement are important elements for the management of BAPU, the most effective measure is the prevention of pressure ulcers through early detection of patients at risk and the adoption of appropriate preventive measures [38‐40].

In conclusion, BAPU is an uncommon complication of pressure ulcers and is mainly caused by S. aureus, Proteus spp., and Bacteroides spp. Its incidence in our institution has declined in recent years. Mortality rate is high and hospital-acquired bacteremia, polymicrobial bacteremia, and serum albumin < 23 g/L are associated with increased mortality.