This site is intended for health professionals only

Clinical interventions:implementing a database


María-José Martínez-Bengoechea
Hospital Pharmacist/Chief Pharmacist

Unax Lertxundi-Etxebarria

Javier Peral-Aguirregoitia

Amaia Santos-Ibáñez

Estibaliz Franco-Lamela

Oihana Mora-Atorrasagasti
Servicio de Farmacia
Hospital de Galdakao
Galdakao (Vizcaya)
E:[email protected]

Computerised physician order entry (CPOE) is being promoted to improve prescribing and ­transcribing accuracy.(1) But high rates of adverse drug events have been reported in highly computerised hospitals.(2–4) Decision support systems integrated in a CPOE can improve therapy appropriateness.(2) Nonetheless, Van der Sijs et al found that drug safety alerts are overridden by clinicians in 49–96% of cases.(5) Clinical pharmacists have a great impact on the quality and safety of medication use. A CPOE has not yet been adopted in our institution. Instead, pharmacists enter prescriptions into the hospital computer and use the information for clinical pharmacy purposes. We describe a semi-automated method to search for drug-related problems (DRPs) in all the inpatients of a 400-bed general hospital.

On a daily basis the pharmacy receives the ­following files about our inpatients via email or intranet: prescriptions, laboratory values, demographics, next-day surgical procedures that require ­antimicrobial prophylaxis or postoperative pain protocols, allergies and pneumoccocal vaccine candidates. Excel(®) or Access(®) macros (elaborated by staff pharmacists) estimate creatinine clearances (CrCl), Child–Pugh scores and digoxin levels. Other queries search for clinically relevant drug interactions, detect serum creatinine increases while in hospital and compute length of ­antibiotic therapy. In addition, some modules update data on allergic patients, look for adverse drug reaction (ADR)-alerting drugs and add daily statistics to a historic file. Useful information about patients, drawn from computerised nurses’ notes, microbiological cultures and previous clinical episodes must be re-entered manually in our database.

Old method: DRP-oriented
Afternoon-shift pharmacists (4–8pm) print out lists:

  • Patients having CrCl <50ml/min and drugs that require dose adjustment in renal impairment.
  • Patients with hyper- or hypokalaemia and drugs that alter potassium levels.
  • Patients on digoxin and estimated digoxin plasma levels.
  • Allergies and projected antibiotic prophylaxis and postoperative analgesics for patients who will be operated on the next day.
  • Drug-allergic patients and their treatments.
  • ADR-alerting drugs.
  • Interactions.
  • Drugs marked for close scrutiny.
  • Personalised letters to patients who are candidates for pneumoccocal vaccination,in case they want to ask their doctor for a vaccine prescription.

After reviewing the lists, the pharmacist marks the problematic drugs in the database. The program sends the corresponding patient’s demographics, laboratory values and drug to the interventions file.

Letters to the GPs are automatically printed. Preprinted labels for renal impairment dosing are available. However, nonstandard advices must be ­handwritten. Letters are sent to the ward by pneumatic tube.

New method: patient-oriented
All the above-mentioned data are combined in a single chart classified by patient. To facilitate follow-up of interventions results, previous pharmacists’ interventions and notes about patients are also included. Pharmacological treatments are classified by ATC (Anatomical Therapeutic Chemical), which helps to detect therapeutic duplications. Symbols appear next to drugs that require renal (R/), hepatic (H/) or elderly (A/) dose adjustments, alter potassium levels (K/) or require special attention (*) (see Figure 1). Early in the morning, each clinical pharmacist receives a list of his/her patients to check for DRP. It is the pharmacist’s decision whether to intervene by telephone and/or by written note. The 4th-year pharmacy resident uses this information on rounds with doctors. The register of pharmacists’ interventions is the same as before.


During the year 2005, there were 16,396 admissions and 122,243 patient-days. Patients were 62±19 years old. Pharmacists reviewed 2,621±440 prescription lines daily, corresponding to 315±51 patients per day. An elevated proportion of patients (41.9%±4.7%) had estimated ClCr<50ml/min. The mean age of patients requiring intervention for DRP was 74±15.

The main pharmacists’ interventions are shown in Table 1. Positive responses from physicians rose from 33% with the DRP method to 61% after the new patient-oriented system was implemented (see Table 2). In addition, there were 189 alerts about drugs to which patients were allergic. The offending drug was withdrawn in 49% of cases. The ­remaining prescriptions continued because the allergy had either been discarded or because it was questionable. Pharmacists evaluated 353 interactions that were in theory clinically relevant, but only alerted for 84 (23.8%) of them. Twenty-one percent out of 1,254 patients who received a letter from the Pharmacy asked for pneumoccocal vaccination. Ninety-seven yellow cards were sent to the Basque Centre for Pharmacovigilance. At present, DRP review requires two fulltime pharmacists’ equivalents: three hours × three pharmacists in the morning and four hours × one pharmacist in the afternoon. The semi-automated method has allowed us to follow up systematically on three new DRPs: drugs in the elderly, interactions and hepatic impairment.




  1. Colpaert K, Claus B, Somers A, et al. Impact of ­computerized ­physician order entry on medication ­prescription errors in the intensive care unit: a controlled cross-sectional trial. Crit Care 2006;10:R21.
  2. Koppel R, Metlay JP, Cohen A, et al. Role of ­computerized physician order entry systems in facilitating ­medication errors. JAMA 2005;293:1197-203.
  3. Han YY, Carcillo JA, Venkataraman ST, et al. Unexpected increased mortality after implementation of a commercially sold computerized physician order
  4. entry system. Pediatrics 2005;116:1506-12.
  5. Nebeker JR, Hoffman JM, Weir CR, et al. High rates of adverse drug events in a highly ­computerized ­hospital. Arch Intern Med 2005;165:1111-6. Van der Sijs H, Aarts J. AU, Vulto A, et al. Overriding of drug safety alerts in computerized ­physician order entry. J Am Med Inform Assoc 2006;13:138-47.

Be in the know
Subscribe to Hospital Pharmacy Europe newsletter and magazine