Pre-emptive isolation involves measures designed to avoid contact with patients strongly suspected to be colonized or infected by multiresistant bacteria (MRB).1 The first references to pre-emptive isolation correspond to epidemic or endemic episodes due to methycillin-resistant Staphylococcus aureus (MRSA), where the admission to hospital of patients colonized or infected by this pathogen was shown to be the origin of a subsequent epidemic.2,3 Some authors justify the adoption of pre-emptive isolation measures in high risk patients, and these measures combined with other interventions designed to prevent cross-transmission have been useful in reducing the incidence of new colonization or infection by this pathogen.4–7

Departments of Intensive Care Medicine or Intensive Care Units (ICUs) have a high risk of cross-transmission of MRB, due to the characteristics of the admitted patients, the many control or treatment interventions made, and the use of complex apparatuses that can act as transmission vectors or reservoirs. As a result, some ICUs have proposed the adoption of pre-emptive contact isolation measures upon the admission of patients at a high risk of being colonized by MRB.8,9

The criteria for indicating such isolation have not been well established, and consensus is likewise lacking regarding the samples that must be collected to study the presence of MRB, and the criteria for suspending isolation.10 In our hospital, pre-emptive isolation of critical patients has been indicated for years, in the absence of any pre-established protocol. This has resulted in different care problems (absence of initial vigilance samples, inadequate prolongation of isolation, unjustified indications of isolation, etc.). With the aim of optimizing our pre-emptive isolation strategy, a consensus-based protocol for pre-emptive isolation (CPPI) was developed for application at the time of patient admission to the ICU. This protocol included the criteria or reasons for isolation, the mandatory samples to be collected for study, the criteria for suspending isolation, and the measures to be adopted in those patients in which isolation is indicated. The present study reports the clinical impact of the application of this protocol in the multidisciplinary ICU of a general hospital.

Of the 1740 patients admitted to the ICU (1055 in the first study period and 685 in the second), pre-emptive isolation was indicated in 199 cases (11.4%). A total of 111 of these patients (10.5%) corresponded to the historical phase (control group) and 88 (12.8%) to the phase following introduction of the CPPI (intervention group). The mean age (±SD) of the patients in which isolation was indicated was 64.5±13.8 years; 109 were males (54.8%); and severity as rated by the APACHE II score upon admission to the ICU was 16.8±8.1. There was a predominance of medical cases (n=183; 92.0%). Nineteen of the patients (9.5%) had been previously admitted to intensive care. The mean duration of stay in the ICU was 11.9±15.5 days, versus a mean hospital stay of 37.6±28.8 days. Thirty-two patients (16.1%) died in the ICU, and another 25 patients (12.6%) died in hospital after discharge from the ICU. The cumulative in-hospital mortality rate was 28.6%. Table 1 shows the characteristics of the study groups. There were no differences between groups in terms of age, background disease, severity score or comorbidities. However, the male gender was seen to predominate in the intervention group (70.5% versus 42.3% in the control group; p<0.001).

The reasons for indicating pre-emptive isolation are reported in Table 2, differentiating between the period before and after introduction of the CPPI. In both periods the predominant indication was admission to a hospital center during more than 7 days in the last 6 months (including current admission) and the use of antibiotics during 7 or more days in the 15 days prior to admission to the ICU. The rest of the indications were much less frequent. The second period was characterized by a significant increase in the indication of isolation due to the presence of chronic ulcers or wounds (1.8% versus 6.8%; p=0.024) and a significant decrease in indications without clinical criterion for isolation (29.7% versus 6.8%; p<0.001). The “correctly indicated isolation” rate was seen to increase following introduction of the CPPI, from 70.3% to 93.2% (p<0.001).

One or more samples for vigilance study were collected in 189 patients (95% of the total). Rectal samples predominated (95% of the patients), along with pharyngeal exudates (49.7%) and blood samples (67.8%). Following introduction of the CPPI in the second study period, a significant increase was recorded in the number of oropharyngeal samples (18.9% versus 88.6%; p<0.001), with a decrease in urine (46.8% versus 25.0%; p=0.002) and blood samples (73.0% versus 61.4%; p=0.050). In turn, the time elapsed between admission to the ICU and sample collection decreased significantly between the two study periods, from 1.56±2.0 days in the control group to 0.37±0.6 days in the intervention group (p<0.001). Likewise, the duration of pre-emptive isolation decreased from 4.77±2.0 days to 3.58±1.4 days (p<0.001). Table 3 shows the samples obtained in each period, as well as their yield in identifying MRB. The rectal samples were seen to be more effective in identifying MRB in both study periods (16.2 versus 17.0%; p=0.763), while the effectiveness of the rest of the samples was very low (<5%).

One or more pathogens were identified in the vigilance samples or in the samples obtained upon admission in 135 of the patients subjected to pre-emptive isolation (67.8%). In turn, 56 MRB were identified in 44 of these patients (22.1%)–their distribution in the two study periods being shown in Table 4. The “adequate isolation” rate was 19.8% in the period before application of the CPPI (control group) and 25% in the period after application of the CPPI (intervention group) (p=0.382). The “adequate isolation” rates corresponding to the different criteria used to indicate pre-emptive isolation are shown in Table 5. Of note is the use of antibiotics during more than 7 days in the 15 days before admission to the ICU (29.3%), a history of colonization or infection by some MRB (27.3%), and admission to a hospital center in the last 6 months (24%).

The application of a consensus-based protocol for pre-emptive isolation (CPPI) in the multidisciplinary ICU of a general hospital has been found to be associated to a significant increase in correct patient isolations and to a decrease in delay in obtaining samples at patient admission to intensive care, as well as a shortening of pre-emptive isolations in those patients who did not need isolation measures. In contrast, there was no significant increase in the adequate isolation rate, which remained in the range of 25% after introduction of the CPPI, thereby meaning that four pre-emptive isolations would have to be indicated in order to avoid one risk of cross-transmission in a patient with MRB.

Although different clinical and microbiological criteria have been proposed for indicating pre-emptive isolation of patients admitted to hospital,1,11,12 no specific criteria have been developed in the case of patients requiring admission to the ICU. In our study, the most frequent reason for indicating pre-emptive isolation was admission to a hospital center during more than 7 days in the 6 months before admission to the ICU. This definition includes patients admitted to hospital during more than 7 days before admission to the ICU. On the other hand, the use of antimicrobials during more than 7 days in the 15 days before admission was also used as criterion for pre-emptive patient isolation. In many cases both criteria coincide in one same patient. The yield or effectiveness of these two criteria in identifying patients with MRB is in the range of 25%. Both criteria have been previously identified in multivariate analyses designed to explore risk factors for methicillin-resistant Staphylococcus aureus carrier status among patients upon admission to hospital13–17 or the ICU.18 These criteria, with some modifications, were proposed by the American Thoracic Society and the Infectious Disease Society of America in screening for patients at risk of presenting MRB as a cause of healthcare-related pneumonia.19 Posteriorly, Nseir et al. analyzed the capacity of these criteria to predict MRB colonization or infection upon admission to the ICU,20 and identified the use of antibiotic treatment in the three days before admission to the ICU, previous hospitalization during two days or more in the three months before admission to the ICU, and patient referral from sociosanitary care centers as independent factors associated to MRB upon admission to the ICU.

The application of contact precaution measures in hospitalized patients is not without risks, and the medical literature has associated such measures to a decrease in patient care by the healthcare staff, an increase in patient depression and anxiety, lesser satisfaction with the care received, and an increase in recorded adverse effects.21–25 A study of adverse effects in the ICU has shown isolated patients to suffer more such effects, and particularly a larger number of errors in prescribing anticoagulants, as well as a greater frequency of detection of hypo- and hyperglycemia.26 The results of our study indicate the need to establish more specific criteria in order to optimize the indication of this type of isolation in critical patients, and to avoid needless isolation.

In the present study, la identification of MRB was made fundamentally in rectal samples, and less frequently in oropharyngeal samples. The identification of MRB in clinical samples has been very low, which justifies the active search for MRB upon admission to the ICU through the collection of vigilance samples. At present, it is debated whether such sampling should be limited to patients at risk of colonization or should be performed on a universal basis.26,27

The creation of mixed working groups headed by nursing staff is an important development for improving the application of pre-emptive isolation strategies.28 Our study shows the application of a CPPI headed by nursing staff to be associated to a greater number of correctly indicated isolations, a greater number of recommended samplings and a decrease in unnecessary samplings, and a shortening of the time elapsed between patient admission to the ICU and the collection of samples for vigilance studies. All this has been possible thanks to the active participation of the ICU nursing staff, which has carried the responsibility of managing the identification of isolation criteria and the collection of vigilance samples. Capacitation of the nursing staff in turn has been completed with the responsibility of complying (and ensuring compliance) with the pre-emptive isolation norms on the part of all the healthcare staff members that assist the patients and their families.

Our study has shown that application of the CPPI results in a significant decrease in pre-emptive isolation time. This is attributable to shortening of the interval between patient admission to the ICU and vigilance sample collection and processing, as well as to the decision to suspend isolation in the case of patients for which the sample results had not been received from the Department of Microbiology by day 5 of admission. This decision was made on examining the results of the vigilance samples corresponding to the control group (i.e., the pre-CPPI period), where it was seen that the MRB-positive cases were reported within the first 5 days after sending the samples to the laboratory. The risk of MRB in those samples without reported findings after 5 days was zero.

The main limitation of this study is represented by the fact that it is a single-center investigation, and the conclusions drawn are only applicable to that setting. Likewise, part of the information was obtained on a retrospective basis, and some of the variables entered in the CRFs were not found in the case histories.

In conclusion, our study for the first time proposes pre-emptive isolation criteria referred to a polyvalent ICU, with the creation of multidisciplinary working groups in order to optimize their application with the purpose of minimizing the spread of MRB in the ICU. The clinical impact of the CPPI has been evidenced by an increase in the “correctly indicated isolation” rate and a decrease in delay in sample collection, unnecessary isolation days, and the duration of indicated patient isolation – thereby reducing the total days of pre-emptive isolation. However, in reference to critical patients, we need new studies or the introduction of new rapid MRB identification techniques allowing the early identification of MRB carries and the avoidance of unnecessary pre-emptive isolations.

The authors declare that they have no conflicts of interest.