Computerized physician order entry-based system to prevent HBV reactivation in patients treated with biologic agents: The PRESCRIB project


  • Blanca Sampedro,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Gastroenterology Department, Galdakao Hospital, Bizkaia, Spain
    Search for more papers by this author
  • Cándido Hernández-López,

    1. Department of Pharmacology, Therapeutics and Toxicology, Taulí Parc Unit, Autonomous University of Barcelona, Sabadell, Barcelona, Spain
    2. Medical Department, Gilead Sciences, Madrid, Spain
    Search for more papers by this author
  • José Ramón Ferrandiz,

    1. Hospital Pharmacy Service, Marqués de Valdecilla University Hospital, Spain
    Search for more papers by this author
  • Aitziber Illaro,

    1. Hospital Pharmacy Service, Marqués de Valdecilla University Hospital, Spain
    Search for more papers by this author
  • Emilio Fábrega,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    Search for more papers by this author
  • Antonio Cuadrado,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    Search for more papers by this author
  • Paula Iruzubieta,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    Search for more papers by this author
  • Susana Menéndez,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    Search for more papers by this author
  • Joaquín Cabezas,

    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    Search for more papers by this author
  • Javier Crespo

    Corresponding author
    1. Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Spain
    2. Marqués de Valdecilla Research Institute (IDIVAL), Santander, Spain
    3. Department of Medicine and Psychiatry, School of Medicine, University of Cantabria, Cantabria, Spain
    • Address reprint requests to: Dr. Javier Crespo, Gastroenterology and Hepatology Unit, Marqués de Valdecilla University Hospital, Avda. Valdecilla s/n, E-39008 Spain. E-mail:; fax: +34 942 202544.

    Search for more papers by this author

  • The PRESCRIB Project received an unrestricted financial grant from the Medical Affairs Department of Gilead Sciences Spain. This financial support consisted of the subroutine programming within the electronic prescription system of the hospital which allows the generation of alerts, requests for analysis, etc.

  • Potential conflict of interest: Dr. Crespo has received consultancy and speaker fees from Bristol-Myers Squibb, Gilead, Roche, MSD, and Janssen. Dr. Hernández-López is currently an employee at the Medical Affairs Department of Gilead Sciences in Spain and owns stock in Gilead.


Computerized physician order entry (CPOE) applications are widely used to prevent medical errors. In our center, a CPOE system has been in use since 2009 on both the inpatient and outpatient levels. A new and simple alert was introduced in the CPOE system to notify healthcare providers of the potential risk of viral reactivation when prescribing biological therapies, thereby facilitating the request for a serological profile (hepatitis B surface antigen [HBsAg], anti-HBc, and anti-HBs) in patients who have not had these tests. Between May 2012 and May 2013, a total of 1,076 patients undergoing biological treatment were included in the implementation of the CPOE in our hospital, resulting in the identification of 4 HBsAg-positive and 69 anti-HBc-positive/HBsAg-negative patients, two of them with positive viral loads. Since the implementation of this alert system, over 90% of patients who were prescribed a biological drug (BD) have undergone serological screening to detect hepatitis B virus (HBV) infection. The use of the alert system has increased the screening rate from less than 50% to 94% for HBsAg and from less than 30% to 85% for anti-HBc in patients for whom a BD is prescribed. Six patients received prophylactic antiviral therapy. No patient had HBV reactivation. Conclusion: This study demonstrates the feasibility of implementing a CPOE system that has allowed our hospital to increase the rate of HBV screening. Its use has facilitated the identification of patients at high risk for HBV reactivation and permitted physicians to prescribe prophylactic measures according to current guidelines. (Hepatology 2014;106–113)


American Association for the Study of the Liver Diseases


Spanish Association for the Study of the Liver


biological drug


computerized physician order entry


European Association for the Study of the Liver


hepatitis B virus reactivation


liver function test

Hepatitis B virus reactivation (HBVr) is well documented in patients with resolved infection and inactive HBV carriers receiving immunosuppressive therapy and/or antineoplastic chemotherapy, especially the drug rituximab.[1, 2] The clinical presentation of HBV reactivation is highly variable and may progress asymptomatically without liver function abnormalities but eventually leading to fulminating hepatitis and even death.[3]

HBVr has become especially important in recent years with the advent of more potent chemotherapeutic agents for cancer, the use of biological medical products in multiple autoimmune diseases, and the possibility of prevention when appropriate therapy is instituted at the right time. The risk of reactivation is not uniform in all patients; it depends primarily on the time course of HBV infection, the type of disease, and the type of immunosuppressive therapy used, and it is particularly high in hepatitis B surface antigen (HBsAg)-positive patients receiving rituximab.[3, 4]

Because of this increased risk, various recommendations and guidelines have been proposed. Some guidelines advise screening for HBV serological markers in all patients about to receive immunosuppressive therapy,[5] while others recommend screening of only those subjects who are considered to be at high risk for reactivation.[6, 7] The most recent guidelines by the American Association for the Study of Liver Diseases (AASLD), the European Association for the Study of the Liver (EASL), and the Spanish Association for the Study of the Liver (AEEH) recommend screening for HBV (HBsAg and anti-HBc) in all patients about to start chemotherapy and immunosuppression. All three guidelines agree that HBV or HBV DNA viral loads should be determined in patients with positive HBsAg or anti-HBc.[8-10]

The PRESCRIB Project was prepared after reviewing our center's global screening results in the last quarter of 2011. The aim of the PRESCRIB project was to create a simple and useful tool to facilitate the screening of all patients who were to be treated with biological drugs (BDs) by medical specialists, to improve interdepartmental communication for the early detection of patients at risk for viral reactivation, and to facilitate early treatment of such cases. The PRESCRIB tool is based on an electronic prescription system (computerized physician order entry [CPOE]). To the authors' knowledge, this is the first published study to explore implementation of a CPOE system in the field of prevention of HBV reactivation.

Patients and Methods

Phases of the PRESCRIB Project

Before starting the study, a preliminary analysis was undertaken to establish our center's baseline screening rates. This analysis was divided into two parts: the purpose of the first part was to characterize the healthcare providers' actual knowledge and routine practices with a survey among 46 physicians; the purpose of the second part was to characterize the screening data in the major specialties prescribing BDs (Rheumatology, Hematology, Oncology, and Gastroenterology) based on the patients' records.

After this preliminary analysis, the project was structured into three phases (Fig. 1). The first phase included seminars on the importance of the risk of HBVr in the context of immunosuppressive therapies as well as the creation of the new CPOE application. After that, there were two sequential phases: an implementation phase and a universalization phase for the new application. In the implementation phase, which took place from May 2012 to November 2012, participation was voluntary. In the universalization phase the participants were recruited with targeted screening conducted by the Hepatology Unit. Subsequently, all BD prescriptions issued between November 2012 and May 2013 were included in the PRESCRIB project. During the implementation and universalization phases of the program, information was collected from all patients who were prescribed a BD in the hospital, in order to evaluate the degree of adherence to the program.

Figure 1.

Phases of the PRESCRIB Project.

Study Design

This was a prospective, observational study to evaluate the usefulness of the CPOE system in identifying patients undergoing biological therapies in our hospital and those at risk for viral reactivation before initiation of treatment with a BD.

A new electronic alert application was implemented as part of the CPOE system previously established in our hospital. This system allows for the introduction of an alert of the risk for viral reactivation associated with a BD at the time the drug is prescribed. At the same time, the prescriber is prompted to enter the patient's serological status. Alternatively, the program itself generates a test order for the patient's serological profile (HBsAg, anti-HBc, and anti-HBs).

A decision-making algorithm was previously established, including: 1) identification of patients at risk for viral reactivation who were to be included in the study (this group included all patients who were initiating treatment with BDs); 2) determination of the best time to introduce the new application alerts on the risk of viral reactivation at baseline; and 3) use of the following alert, which included a reminder of the risk of viral reactivation and a question about the patient's serology: “Dear Dr. — As you already know, the drug that you are about to prescribe can potentially cause the reactivation of HBV. Do you know this patient's serology?” Three possible answers could be given to this question (Fig. 2): 1) Yes, it is positive (HBsAg positive or/and anti-HBc positive); 2) Yes, it is negative (HBsAg negative, anti-HBc negative); and 3) No. If the response was that the patient's serology was unknown, the application itself automatically generated a patient serology order (HBsAg, anti-HBc, anti-HBs).

Figure 2.

Algorithm followed by the CPOE application for HBV screening.

The patients' serological data were collected by the Hepatology Unit staff, who decided whether patients required follow-up by the Hepatology Unit. These staff members were responsible for confirming adherence to the program.

Patients positive for HBsAg and/or anti-HBc were followed by the Hepatology Unit by determination of serum HBV DNA every 3 months. In these patients, antiviral treatment was prescribed following the recommendations of the AASLD and EASL guidelines.[8, 9] In cases where completeness of serologic information could not be verified, the patients were excluded from the final statistical analyses.

The study was approved by the Clinical Research Ethics Committee of Cantabria and was conducted according to Good Clinical Practice guidelines and the Declaration of Helsinki.

Patient Selection and Inclusion Criteria

Between May 2012 and May 2013, we evaluated 1,076 patients. Patients were included in the study if they met the following criteria: 1) treatment with BDs during the study period; 2) agreement to participate in the study; 3) available medical history and physical examination; and 4) available HBV infections markers.

Data Collection

Serological data for HBV were collected, including tests for the presence of HBsAg, anti-HBs, and anti-HBc antibodies (HBsAg: ARCHITECT HBsAg QT; anti-HBs: ARCHITECT Anti-HBs; anti-HBc: ARCHITECT Anti-HBc, Abbott Laboratories, Wiesbaden, Germany). Data were collected from all patients enrolled in the study and included information regarding gender, age, ethnicity, underlying disease, referring department at the Marqués de Valdecilla University Hospital, hepatic viral serology (see above), and liver function test (LFT) results (transaminases and bilirubin). In patients who were positive for HBsAg or anti-HBc with or without anti-HBs, HBV DNA blood levels were measured using real-time polymerase chain reaction (PCR). HBV DNA levels were quantified using the automatic COBAS TaqMan HBV Test 2.0 equipment, with 15 IU/mL sensitivity.

Definitions Used in This Study

HBV reactivation was defined as an abrupt increase in viral load (≥1 log10 IU/mL) and alanine aminotransferase levels[11] in a patient with inactive or occult HBV infection. HBV occult infection[12] was defined as the presence of persistent viral DNA in liver tissue and/or circulation at very low (generally < 200 IU/mL) or undetectable levels in an HBsAg-negative patient. In cases of occult infection, anti-HBc with or without anti-HBs is usually present, although in exceptional cases all serological markers can be negative.[12] The definitions of inactive chronic carrier and chronic hepatitis B were those universally accepted.[8]

Statistical Analysis

The descriptive analysis of qualitative variables included frequencies. Means and standard deviations were analyzed as quantitative variables. Qualitative variables were measured as medians and ranges.


Patient Characteristics

From May 2012 to May 2013, a total of 1,076 patients undergoing treatment with BDs were reviewed using the PRESCRIB application. Of these patients, 83 were excluded because the serological status could not be confirmed in the hospital records; 993 patients were included in the final analysis; in 60 of them the serology was requested by the program but never obtained. There were 425 males and 568 females, with median ages of 54 (6-89 years) and 56 (2-92 years), respectively. Only 2% of the population came from countries with high prevalence rates of HBV infection. The implementation phase included 487 patients, whereas 506 patients were included in the universalization phase. Figure 3 shows the flow of patients. The patients' demographic and clinical characteristics are shown in Table 1.

Table 1. General Patient Characteristics
Gender (male/female)425/568
Age in years, median (range)
Males54 (6-89)
Females56 (2-92)
Patients by origin
Western Europe970
Eastern Europe13
Patients by specialty involved (%)
Rheumatology361 (36.3)
Medical Oncology244 (24.5)
Hematology141 (14.1)
Dermatology96 (9.6)
Gastroenterology (No hepatologists)83 (8.6)
Other (Neurologists, Internal Medicine, etc)68 (6.8)
Biological drug used (patients)
Figure 3.

Patient disposition during the CPOE application and according to the risk of HBV reactivation.

The type of BD and the prescribing department are shown in Table 1. Many of the patients in the study were referred from the Rheumatology (36.3%) and Medical Oncology (24.5%) departments, and a relatively large percentage of the patients had oncohematologic diseases (14.2%). The most common BDs reported were rituximab, adalimumab, and infliximab.

HBV Screening at Baseline

The physician survey was carried out as part of a national survey that included 19 hospitals in our country.[13, 14] In our hospital, 70% of the participants were from the Gastroenterology, Rheumatology, or Hematology departments, overall reporting 69% and 63% screening rates for HBsAg and anti-HBc, respectively, prior to starting immunosuppressive treatments. The screening rate among Oncology specialists, around 20%, was the lowest (Table 2).

Table 2. Screening (HBsAg and Anti-HBc) Rates According to the Phase of the PRESCRIB Project and the Medical Specialty Involved
Physician SurveyPatient ScreeningImplementation PhaseUniversal-Screening Phase
  1. CPOE: Computerized physician order entry; NA: No data available.

  2. a

    Patients during implementation phase are also included in the universalization phase; screening failures not included; when included, the overall rate of HBsAg and anti-HBc screening was 87% and 78%, respectively (see text and Fig. 3).

Specialty% Physicians HBsAg-anti-HBc (n)% Patients HBsAg-anti-HBc (n)
Rheumatology50-25 (4)57-29 (328)80-53 (148)96-84 (361)
Oncology10-20 (10)10-2 (233)46-24 (132)89-86 (244)
Hematology100-89 (10)63-48 (139)86-66 (77)94-90 (141)
DermatologyNANA94-49 (48)94-72 (96)
Gastroenterology89-83 (19)85-80 (80)93-77 (42)100-98 (83)
Other67-67 (3)NA93-44 (40)88-79 (68)
TOTAL69-64 (46)47-29 (780)75-48 (487)94-85 (993)a

The actual HBsAg and anti-HBc screening rates for the same period as determined from the patients' records showed a significantly lower rate of HBV screening (47% for HBsAg), mainly for anti-HBc (29%) and Oncologists (2%) (Table 2).

HBV Screening During the Universalization Phase of CPOE

As shown in Table 3, HBsAg status was known at BD initiation for 94% (933/993) of the study patients; 85.1% (845/993) had anti-HBc serology and 84.8% (842/993) had anti-HBs serology (all of them had been tested for HBsAg). The prevalence of positive HBsAg, anti-HBc, and anti-HBs was 0.4%, 8.6%, and 16%, respectively. In all four HBsAg-positive patients, HBV DNA was detected. All four of these patients were treated with an antiviral high genetic barrier drug, which resulted in the steady decline and clearance of the viral load. No reactivation of HBV infection in these patients was documented during follow-up. In 69 HBsAg-negative patients (8.2%), the presence of positive anti-HBc was documented. The characteristics of these patients are shown in Table 4. Notably, 62.3% did not show LFT abnormalities. Fifty-eight were also anti-HBs-positive, with an average titer of 268.32 (±361.65) IU/mL. Eleven patients had an isolated anti-HBc-positive result. The median follow-up of HBsAg-positive patients was 11 (9-15 months), 13 (6-21 months) for the anti-HBs and anti-HBc-positive patients, and 11 (6-16 months) for occult HBV-infected patients. Only two anti-HBc-positive patients had a positive viral load under 200 IU/mL; both were HBsAg-negative, and one had anti-HBs >1,000 IU/mL. Both of these patients received high genetic barrier antiviral therapy (tenofovir DF) and were negative for viral load at the next test. No reactivation of HBV infection in these patients was documented during follow-up.

Table 3. Serological Parameters of the Patients
n = 993KnownUnknown
  1. Results expressed as number of patients (%).

  2. a

    Negative HBsAg.

  3. b

    All patients screened for anti-HBc and anti-HBs had also been screened for HBsAg.

HBsAg4 (0.4)929 (99.6)60 (6.04)
Anti-HBcb73 (8.6)772 (91.4)148 (14.9)
-Inactive carriers4 (0.4)  
-Occult HBV infectiona69 (8.2)  
Anti-HBsb136 (16.1)706 (83.9)151 (15.2)
Table 4. Characteristics of Anti-HBc-Positive Patients With Negative HBsAg (n = 69)
  1. LFT: liver function test; ULN: upper limit of normal.

Males/Females (n)26 / 43
Age, years (mean ± SD)
Males62.9 ± 15.6
Females61.4 ± 13.1
Underlying disease (n)
Serology (n)
Isolated anti-HBc (+)11
Anti-HBc (+) and anti-HBs (+)58
HBV DNA-positive2
LFT abnormalities (n)
< 2 × ULN14
> 2 × ULN12
LFT parameters (mean ± SD)
AST (U/L)33.82 ± 45.68
ALT (U/L)34.4 ± 34.70
Bilirubin (mg/dL)0.83 ± 1.96

A subgroup of great significance, given the high risk of viral reactivation, includes rituximab-treated patients. In these patients the screening rates were 93% (170/182) for HBsAg and 87% (159/182) for anti-HBc; 129 were oncohematologic patients, 45 had inflammatory diseases, and eight were renal transplant recipients. The patients within this group who were at increased risk for viral reactivation included two HBsAg-positive patients and 25 with positive anti-HBc.

CPOE Impact: Evolution of the HBV Screening Rates in Patients

The evolution of the HBV screening rates in our hospital is shown in Table 2. Regarding HBsAg screening, the baseline rate was below 50% and there was an increase to 75% during the implementation phase, eventually reaching 94% in the universalization phase. The baseline rate for the anti-HBc parameter was lower (29%), increasing to 48% during the implementation phase and achieving 85% at the end of the phase. The increase in the screening rate was achieved in all the departments involved for any type of prescription and under all conditions tested.

There were 231 patient screening failures in this study (Fig. 3): 21% of all 1,076 patients included in the CPOE project. The majority of these program failures (127) occurred during the implementation phase. Counting all screening failures, a more conservative estimate of overall screening rates would be 87% and 78% for HBsAg and anti-HBc, respectively.


The adoption of the CPOE system in our hospital resulted in the screening of almost all patients treated with BDs, regardless of the BD type, the department responsible for the patient, and the type of underlying disease. The use of this tool allowed the hospital to increase its HBsAg screening rate from below 50% before the implementation of the program to 94% during the universalization phase. The anti-HBc screening rate increased from 29% to 85%. It is important to emphasize that the rates of screening in the implementation phase probably were aided by educational efforts. However, we believe that the final remarkable result was achieved by combining an educational effort, which is insufficient by itself, with the development of a technology such as the CPOE.

Historically, CPOE use has improved patient safety, prevented medical errors,[15, 16] and facilitated communication between physicians and pharmacists.[17] The PRESCRIB Project established an innovative strategy that further develops the CPOE capabilities introduced some years ago by our hospital pharmacy. This strategy involves a protocol of alerts for doctors who prescribe BDs, alerting them about the possibility of reactivation of hepatotropic viruses.

Some potential advantages of CPOE-like systems applied to the HBV screening have been suggested previously. Liu et al. noted that the implementation of CPOE in HBV screening for cancer patients who are to receive rituximab allowed the dispensing pharmacist to confirm this screening and to notify the treating physician if screening had not been performed. However, as suggested herein, the implementation of the CPOE requires staff and adequate training, which we aim to achieve; however, perhaps not all centers may be able to meet this requirement.[18] Furthermore, as some authors have pointed out, the adoption of CPOE has been demonstrated to be slower and more problematic than anticipated, with adoption rates around 20% or less.[19] Nevertheless, the advantages of CPOE remain compelling, as the PRESCRIB Project has demonstrated.

Previous studies have noted that some specialties, such as Oncology, tend to have lower overall rates of HBV screening.[14, 20] The results reported here for the PRESCRIB Project also show differences in screening rates by specialty. Indeed, Gastroenterology and Hematology have screening rates of over 80%, whereas in our hospital the initial screening rates at the Oncology Unit were below 50%. Although this problem was overcome by performing a targeted screening by the Hepatology Unit, it would likely be appropriate to pursue greater awareness in this group of specialists so as to encourage their involvement and ultimately achieve universal screening.[5, 14]

Compared to other publications,[21-23] prior to initiation of our project there was a high prevalence of HBsAg and anti-HBc screening, probably due to the higher percentage of prescriptions by Rheumatology and Hematology specialists in our center compared to other specialties including mainly Oncology specialists. This specialty has shown clearly lower screening rates.

The most frequently used BDs in our study were the tumor necrosis factor alpha (TNF-α) inhibitors infliximab and adalimumab and the anti-CD20 antibody rituximab. With these BDs, the risk of HBVr is variable and described elsewhere.[24-32] What is significant to note is that HBV reactivation may occur in patients with positive HBsAg, negative HBsAg, and positive anti-HBc, and even with the presence of anti-HBs.[33-35]

Anti-HBs antibody titers have been reported to have a protective effect in HBVr.[36, 37] Our data do not confirm these findings, as no reactivation has been detected at this time. Remarkably, one of the patients with positive anti-HBc antibody and positive DNA had an anti-HBs level of over 1,000 IU/mL.

CPOE has enabled our hospital to identify 73 patients with markers of HBV infection (HBsAg and anti-HBc). The 0.4% prevalence rate of positive HBsAg found in this study is slightly lower than the 0.7% rate that was previously documented in Spain.[38, 39] This probably reflects a decreasing trend, from the 1.2% rates reported in the 1980s to 0.7% in 2002[39] and 0.4% nowadays. Regarding the presence of positive anti-HBc with negative HBsAg, CPOE reported a prevalence of 8.2% (69 patients), which is lower than the value of 8.7% reported in 2002.[38, 39] The implementation in our country since 1992 of universal vaccination campaigns in children and newborns may be the reason for this downward trend, especially because immigration is very low in our hospital area.

If reactivation rates as generic and conservative as 25% for HBsAg-positive patients and 5% for positive anti-HBc and negative HBsAg are applied,[24, 31, 32, 39, 40] the use of CPOE would have avoided at least four HBVr cases in our hospital. In fact, in a recent publication on patients with rheumatic diseases it was demonstrated that knowing the patients' serology at the beginning of BD treatment could theoretically prevent up to 78% of reactivation episodes.[41]

It is reasonable to wonder whether universal HBV screening by CPOE is cost-effective, especially under the current conditions of economic hardship. Only two studies have analyzed the cost-effectiveness of screening for HBV infection in patients who are about to receive chemotherapy or immunosuppression. A study of patients with solid tumors showed that such screening was not cost-effective in patients with metastatic disease, a result that was most likely associated with mortality. However, the use of the HBsAg tests in other cancer patients was indeed cost-effective.[42] The cost-effectiveness of universal screening is clearer in the context of lymphomas, in which HBV screening by HBsAg in patients who are about to receive regimens including rituximab was shown to be cost-effective in any area, with a prevalence rate of HBsAg ≥0.2%; this includes the USA, Canada, Australia, and most European countries.[43] However, it remains unclear whether inclusion of the anti-HBc marker is also a cost-effective practice in this situation. A future target of the CPOE use in HBV reactivation should be to achieve equal rates of HBsAg and anti-HBc screening (preferably 100% for both), because the rate of HBsAg screening is currently higher than anti-HBc screening, as shown in our study.

During development of the PRESCRIB Project, six patients received antiviral prophylaxis; four of them were HBsAg-positive, and two had positive anti-HBc and negative HBsAg but detectable viral load. The authors agree with the latest recommendations and guidelines on the management of this infection.[8-10]

In conclusion, the PRESCRIB Project has shown that it is feasible to use CPOE in our hospital to increase HBV screening for patients who are about to receive immunosuppressive drugs. The use of the PRESCRIB system has allowed physicians to identify patients at a higher risk for HBV reactivation and to provide prophylactic measures according to current guidelines. If it is accepted as a paradigm that HBV reactivation can be prevented with prophylactic oral antivirals, it is obvious that preventive screening should be a part of that strategy. Because a tool such as the CPOE can facilitate this preventive screening, it would be interesting to determine whether this experience can be extrapolated to other centers with CPOE systems and whether this is cost-effective for hospitals.


The authors thank Dr José Mora-Maciá who helped prepare the article.