INTRODUCTION
In many developing countries, the availability and the use of antibiotics are poorly controlled, resulting in high rates of microbial resistance1. Brazil has a Unified Health System (SUS), which provides for all actions and health services which are gratuitously provided by the governments. SUS is applied to the entire Brazilian population2, and the health authorities are concerned about the proper use of antibiotics3.
Serious bacterial infections should be, and are traditionally treated with IV antibiotics. However, after clinical improvement, the oral treatment (OT) can be used. Besides, the prolonged and unnecessary use of IV, although not desirable, usually occurs4.
One of the data used to evaluate the use of antibiotics in hospitals described by the European Surveillance of Antimicrobial Consumption (ESAC) is the proportion of oral versus parenteral use. (ESAC moved to European Centre for Disease Prevention and Control – ECDC – in 2011 and is now named ESAC-Net)5.
Sequential antibiotic therapy (SAT) refers to the exchange from the parenteral route to the oral treatment as soon as the patient is clinically stable. Clinical and laboratory criteria are suggested so as to identify the patients who are sufficiently stable to enable therapy change6. It is already clear that SAT is safe, and economical, and that it improves the quality of healthcare4,7.
Oral formulations are cheaper then IV ones, leading to a reduction in the following: time of preparation and administration, work of nursing staff, drug waste and length of hospitalization. The OT is also easy to continue at home8-10. Moreover, the reduction of the hospital stay and the length of catheter use, due to SAT, may lead to a reduction in hospital infection incidence8,11,12.
As a consequence of the increasing economic pressure, it becomes necessary to control hospital costs which are deeply influenced by the use of intravenous drugs, thus the SAT strategy is being more and more implemented. Therefore, studies are considered necessary in order to reduce the use of IV antibiotics in the treatment of hospitalized patients13. Previous studies with SAT differ as to the characteristics of the intervention, of the hospital, of the infectious syndromes and of the antibiotics tested6,14-17.
In this context, the purpose of this present study was to discover the frequency of SAT practice and, especially, an effective strategy to reduce the use of IV antibiotics by the implementation of the SAT in a teaching hospital in Brazil.
METHODS
This prospective, historically controlled, interventional study was conducted at the Clinical Hospital of the Federal University of Uberlândia (HCU). The HCU is a public teaching hospital, having 503 beds and has a 100% health insurance agreement with the Unified Health System (SUS) of Brazil. Data were collected during the pre-intervention (PIP) and the intervention (IP) periods.
During PIP, from April 4 to July 20, 2005, from Monday to Friday, the researchers visited all the patients hospitalized in the internal medicine wards (52 beds), in the surgery wards (128 beds) and in the intensive care unit for adults (15 beds). All the patients who initiated use of IV antibiotics on the day of the visit or on the day before were included in the study. These patients were followed up for at least 60 days, or up to discharge or death.
The use of IV metronidazole (because it can be used as an antiparasitic) and the use of IV cefazolin (because it is commonly indicated prophylactically) did not motivate inclusion in the study. Some patients were hospitalized for more than 60 days after beginning the IV antibiotic, but for the calculation of the average time of hospitalization only 60 days were considered. It was considered to be a new course when a new IV antibiotic was introduced after the patient had remained without an IV antibiotic for more than one day. It was considered SAT when, at any time, an oral antibiotic was introduced to replace some IV antibiotic.
From the second half of 2005 Guidelines for the Use of Antibiotics were elaborated. These contain recommendations for the empirical treatment of community-acquired or hospital-acquired pneumonia, urinary tract infections, intra-abdominal infections and for sepsis in adults. They include the recommendation of the ideal time to perform SAT. The guidelines were written by two infectious disease physicians from the Hospital Infection Control Service (HICS), and by one biochemist pharmacist, directly involved in the research. They were based on specialized scientific literature and on the profile of the susceptibility to microorganisms isolated in the HCU. After that, they were submitted for approval to the Control Committee of Antibiotic (CCA) and to the Hospital Infection Control Committee (HICC) of the HCU.
Between PIP and IP, in addition to the elaboration and approval of the guidelines, other facilitator and educational strategies were developed such as the following: a) standardization of new antibiotics by OT: azithromycin 500mg; clindamycin 300mg; moxifloxacin 400mg; sultamicillin tosylate (ampicillin/sulbactam) 375mg; b) availability of electronic version of the guidelines starting from July 20, 2007, on the main page of HCU intranet, accessible by a single click on the icon; c) presentation of the guidelines to physicians and to medical students working in some evaluated wards; d) printed copies were supplied to the physicians on duty; e) printed copies were fixed to the specific murals in the prescription rooms.
After implementation of these strategies, the IP was started on September 24 and finished on December 20, 2007, when the number of patients was the same of the first period. This strategy was done by convenience and the data collection followed the same methodology for data collection as that of the previous period. The same beds that were visited in PIP were also visited in IP.
The following was also performed, daily fixing of labels (Figure 1), from Monday to Friday, on patients’ prescription suggesting to their physicians to revise the therapy with IV antibiotics.
In the PIP, data were collected by the biochemist pharmacist researcher and a third year medical student, and in IP, similar activities were performed by the researcher and a fourth year nursing student. Any doubts that may have arisen during the data collection were discussed with the infectious disease physician, who is the study coordinator.
Statistical analysis
The StatCalc, module of the EpiInfo 2000, a software distributed by the Center of Disease Control and Prevention was used to analyze the practice of SAT. The Chi-square test with Yates correction was used to evaluate the secondary variables. The Mann-Whitney test through BioStat 3.0, national software of public domain was used to evaluate the difference in the duration of the course of antibiotic treatment and the length of hospital stay from the first antibiotics prescription, in both study periods. Results are reported as statistically significant at p < 0.05.
RESULTS
One hundred and seventeen patients were evaluated in the PIP and 117 in the IP. The characteristics obtained are shown in Table 1.
The IV antibiotics prescribed, in the PIP and in the IP, were ceftriaxone (23.4% and 21.7% of the cases, respectively), cefepime (18.3% and 15.6%), vancomycin (11.7% and 8.7%), levofloxacin, gatifloxacin or ciprofloxacin (12.1% and 10.6%), clindamycin (8.8% and 8%), amikacin or gentamicin (7% and 2.7 %), imipenem or meropenem (6.6% and 8%), oxacillin (4.8% and 3%), ampicillin or ampicillin-sulbactam (4% and 7.2%), cefazolin (1.5% and 3%), crystalline penicillin (1.1% and 0.8%), piperacillin-tazobactam (0.4% and 0%) and trimethoprim + sulfamethoxazole (0.4 and 0.4%). In the PIP and IP, respectively, the mean duration of the use of IV antibiotics per course of treatment, was 14.79 and 11.75 days (p < 0.01) and medians were 13.0 and 9.0 days.
In the PIP and PI, 69 (59%) and 66 (56.4%) patients, respectively, received more than one IV antibiotic, simultaneously. Twenty nine (24.8%) and 38 (32.5%) received antibiotics through a central venous catheter (CVC). At the beginning of treatment, 45 (38.5%) and 58 (49.6%) patients, respectively, were not receiving nutrition through the digestive tract, 60 (51.3%) and 46 (39.3%) were nourished orally, 12 (10.3%) and 13 (11.1%) through probe (nasoenteral, nasogastric, oroenteral or gastrostomy) (Table 2).
From the first prescription of antibiotics, the length of the hospital stay was more than 60 days for 6 (5.1%) patients in the PIP and for 5 (4.3%) patients in IP (Table 1). In the PIP and IP, respectively, 96 (82.1%) and 90 (76.9%) patients were discharged, and 15 (12.8%) and 22 (18.8%) died.
In the PIP, SAT occurred in 4 (3%) of the 135 courses of treatment, and in IP, it occurred in 5 (3.9%) of the 130 courses (p = 0.95) (Table 3). It was also observed that, after SAT, there was a return to IV route in two of these cases, one in each study period.
DISCUSSION
The studied populations at different periods, had similar demographic and clinical characteristics, with the predominance of intra-abdominal, and respiratory and urinary tracts infections10,15. The predominance of these infections is common to most hospitals10,15. The largest percentage of cases of respiratory infections found in PIP was higher than that in the IP, and this might be related to the period of the year when this study was conducted. Respiratory infections become more frequent in colder and drier weather18,19.
The data of this study, although collected from a single hospital, raise suspicions that in Brazil the SAT is little performed in many hospitals.
In the present study, the eligibility of patients has not been assessed regarding clinical stability criteria. However, it is known that by the third day of treatment, nearly half of the hospitalized patients are already eligible for SAT17,20.
Some studies showed the effectiveness of the use of guidance protocols for SAT promotion. Al-Eidan et al. 21, in the UK, observed that the use of a protocol led to a reduction in the length of time that IV antibiotics were used for patients with community-acquired pneumonia. The authors attributed the success of the program to three main reasons: the system used to assess patient’s risk factors; the predetermined criteria to exchange the antibiotic to OT, when the patient’s condition stabilized or improved and a clear and simple presentation of an algorithm for infection management. Even before the use of the protocol, the average time (geometric mean) for IV antibiotics was 5.7 days, which shows a great discrepancy from the data of this present study. Although the authors evaluated only community-acquired pneumonia, this is not a consistent reason for this discrepancy. McLaughlin et al. 10. in a Glasgow’s hospital, also in the UK, pointed out that the guidelines intervention adapted for the locality, increased substantially the use of SAT. Likewise the study conducted by Al-Eidan et al. 21 shows that the SAT was already performed in most patients, even before the intervention. Sevinç et al. 15, in a study performed in a large university hospital in the Netherlands, noted that a guideline reduced the average duration of intravenous treatment from 6 to 4 days. The reduction time (6 to 4 days) was even shorter than the one of this present study (14.79 to 11.75 days), but the reduction was proportionately much greater.
Although not evaluated for each individual case whether patients were able to receive the antimicrobial agent by OT, the number of cases of patients who underwent SAT was extremely small. Moreover, half of the patients have clinical stability about 3 days after start the antibiotic treatment17,20,22. It is clear from the current literature that there need not be a total resolution of parameters before oral switch can occur, but that the optimal timing has yet to be determined4,23.
There was no explanation for the shorter hospital length of stay in the second period of the study, but it is not due the implementation of SAT because it doesn´t occurred.
In this present study, in the few cases of the use of SAT, a quinolone or a sulfonamide was used, perhaps because these antibiotics are available in both oral and intravenous presentation. They also have a good spectrum for nosocomial germs and they have very good bioavailablility by OT. For quinolones, the serum concentration is similar to that achieved by IV use24. However, in this present study, the quinolones were used in several other courses of treatment in which the SAT was not used. The third generation cephalosporins have been widely used, and currently there are no oral formulations available for these in the market12. This may also be a limitation for SAT. SAT is safe, effective and economical when performed at the appropriate time4,7. So there is no reason to disregard its use in a public hospital of a still developing country like Brazil.
The ability to tolerate and absorb the antibiotic administered orally can be easily assured when the patient is receiving and tolerating nutrition and/or other drugs, by OT24. This shows that gastrointestinal intolerance to the drug may not be the cause of the permanent use of IV antibiotics in these cases.
Thus, once there is a clear reason to use the SAT, an important issue is to know how to act in order to succeed in changing the behavior of physicians. It was noted that the proposed intervention, although it may be useful in some circumstances, was not so in this study. This shows that several factors, which are not clear to the observer, may interfere with the effectiveness of the interventions. A systematic review of strategies for continual medical education suggests that multiple interventions (bundles), performed with leadership and monitoring, are most effective for information transmission25. However, it is very difficult to prove which of these multiple interventions is truly effective.
One alternative would be to add to the guidelines, an explanation for the prolonged use of IV antibiotics by the physicians so that the pharmacist could easily provide them. Forms filled in by physicians justify the permanence of an intravenous antibiotic therapy for more than 48/72h as happens in the studies conducted in university hospitals in Manchester, England17 and in Lausanne, Switzerland26. In this context, the performance of the clinical pharmacist and of the infectious disease physician of the HICS, are important to assess the justifications and to provide continuous feedback to the physicians.
It is concluded that, in the presented conditions, the evaluated intervention strategy is not effective in the sense of promoting the SAT. It is also concluded that in the studied hospital, SAT is rarely performed and that the length of time of IV antibiotic use is very long.