RESEARCH ARTICLE

Drug prescribing in patients with renal impairment optimizedby a computer-based, semi-automated system

Ana Such Dıaz • Javier Saez de la Fuente • Laura Esteva • Ana Marıa Alanon Pardo •

Nelida Barrueco • Concepcion Esteban • Ismael Escobar Rodrıguez

Received: 19 October 2012 / Accepted: 26 August 2013

� Koninklijke Nederlandse Maatschappij ter bevordering der Pharmacie 2013

Abstract Background According to several studies,

despite of the existence of several published guidelines for

dosing adjustments based on renal function, inappropriate

prescribing is a common drug-related problem in inpatient

care. Objective We developed and implemented a system

for drug dosage adjustment integrated into the Hospital

computer provider order entry system. This system allows

pharmacists to identify patients with reduced renal func-

tion, identify medication orders that may require dosage

modifications based on renal function, and generate an alert

with a recommendation of specific dosage adjustment.

Using the Summary of Product Characteristics and two

drug databases (Micromedex 2.0� and Lexicomp�), spe-

cific dosage guidelines for drugs used in patients with renal

impairment were established. Setting A 264-bed tertiary

teaching hospital. Methods We performed a quasi-experi-

mental, one-group, pretest–posttest study to assess the

efficacy of this intervention program. We compared the

differences between the frequency of appropriate orders

pre- and post-test using the McNemar test. Main outcome

measures: the frequency of appropriate orders before the

recommendation (pre-test) and after the recommendation

(post-test). Results Before the intervention, the frequency

of appropriate prescribing based on renal function was

65 %. After the intervention, this frequency was 86 %

(p \ 0.001). The interventions were more frequent in the

emergency department (45 %). The program required

30–45 min of pharmacist time per day. The average

number of patients reviewed daily was 28. This study

found that a computer-based, semi-automated drug-dosage

program for renal failure patients was able to reduce the

number of inappropriate orders due to renal insufficiency.

Keywords Computer-assisted drug therapy �Decision support systems � Medical order entry

systems � Medication errors � Renal insufficiency �Spain

Impact of findings on practice

A computer-based semi-automated alerts system

enables pharmacists to quickly review medication orders

written for renal failure patients. A computer-based semi-

automated alerts system improves drug prescribing in

patients with renal insufficiency, allows pharmacists to

participate more actively in patient care, and it improves

the basic clinical decision support system, avoiding over-

alerting in a cheap and easy way.

Introduction

Chronic kidney disease (CKD) is a significant and costly

worldwide health problem. According to the Spanish

Society of Nephrology’s (SEN) database of patients with

kidney disease, Spain has one of the highest rates of end-

stage CKD, and this problem will increase in the coming

A. Such Dıaz (&) � J. Saez de la Fuente � N. Barrueco �C. Esteban � I. Escobar Rodrıguez

Pharmacy Department, Hospital Universitario Infanta Leonor,

Gran Vıa del Este, 80, 28031 Madrid, Spain

e-mail: ana.such@salud.madrid.org

L. Esteva

Pharmacy Department, Hospital de Torrejon, Torrejon de Ardoz,

Spain

A. M. Alanon Pardo

Pharmacy Department, Hospital Universitario Virgen de las

Nieves, Granada, Spain

123

Int J Clin Pharm

DOI 10.1007/s11096-013-9843-3

years due to the aging population and the increasing

prevalence of diabetes and hypertension [1].

The ERPHOS study, conducted in Spanish hospitals,

found a glomerular filtration rate (GFR) \60 mL/min in

28.4 % of hospitalized patients and\44 mL/min in 13.1 %

of patients. Additionally, 42 % of men and 59 % of women

older than age 80 had a GFR of \60 mL/min [2].

The National Kidney Foundation Kidney Disease Out-

comes Quality Initiative (K/DOQI) defines chronic kidney

disease as the presence of kidney damage or a reduction in

the GFR to \60 mL/min/1.73 m2 for 3 months or longer.

The K/DOQI’s chronic kidney disease staging is based on

GFR [3].

The consensus of the SEN and the Community and

Family Medicine Spanish Society (SemFYC) is to use an

equation to routinely estimate GFR rather than using serum

creatinine levels alone, because the equations take into

account several clinical and demographic parameters and

are considered the best index of GFR in clinical practice.

The same consensus recommends the use of the Modifi-

cation of Diet in Renal Disease (MDRD) formula for

estimation of GFR, and the traditional Cockcroft-Gault

(CG) formula as an alternative (Table 1).

The Cockcroft-Gault formula aims to predict creatinine

clearance from knowledge of serum creatinine, age,

weight, and sex [4]. The MDRD4 formula considers serum

creatinine, age, weight, sex, and race [5, 6].

Many drugs, or their metabolites, are eliminated through

the kidney. CKD can also affect other pharmacokinetic and

pharmacodynamic processes. So modifications in the dos-

age of a variety of drugs are needed in the presence of renal

failure to avoid drug toxicity, ineffective therapy, and

increased costs [6, 7].

Drugs that require dose adjustments in patients with

renal impairment are commonly used in hospitals and can

result in adverse drug events (ADEs).

According to several studies, despite of the existence of

published guidelines for dosing adjustments based on renal

function, inappropriate dosing of renally cleared or

potentially nephrotoxic drugs is a common drug-related

problem in inpatient care [8].

Improvements in compliance with renal-dosing guide-

lines are needed to achieve good prescribing practices in

inpatients with renal impairment. Good prescribing prac-

tices in patients with renal insufficiency are defined as

orders that are appropriate by dose and frequency accord-

ing to renal dosing guidelines.

Computerized provider order entry (CPOE) with clinical

decision support system (CDSS) can improve patient safety

and lower medication-related costs [9], but at present, these

systems are unevenly developed in Spain [10].

Several studies have been developed to assess the effi-

cacy of CPOE with CDSS and/or review of orders by

pharmacists for reducing prescription errors related to drug

dosage in renal insufficiency. The studies have shown

varying results [11–26].

We developed and implemented a system for drug

dosage adjustment that is integrated into the Hospital

CPOE/CDSS system (excluding intensive-care units and

anaesthesiology wards) and allows pharmacists to identify

patients with reduced renal function, identify medication

orders that may require dosage modifications based on

renal function, and generate an alert with a recommenda-

tion of specific dosage adjustments (based on current

dosage guidelines).

We decided to combine automated-alert systems with

manually reviewed patient charts to avoid unnecessary

alerts and allow enhancement of the process.

Aim of the study

Our goals were to develop a semi-automated alert system

for checking doses of medications according to the

patient’s renal function and determine if this intervention

improves drug prescribing practices.

Method

Study site and setting

Our study was conducted in a 264-bed tertiary teaching

hospital (Hospital Universitario Infanta Leonor, Madrid).

The hospital’s information system (Selene�) integrates

an electronic health record (EHR) and CPOE with basic

clinical decision support system (CDSS). This basic CDSS

includes drug-allergy checking, formulary decision sup-

port, and drug–drug interaction checking. CPOE and lab-

oratory data are integrated with other applications such as

the pharmacy information management system (Farma-

tools�). All inpatient orders (except ICU and anaesthesi-

ology ward orders) are entered into Selene� (EHR with

CPOE).

Table 1 Recommended

equations for routine estimation

of GFR4 [4, 5]

GFR glomerular filtration rate

COCKCROFT-GAULT Creatinine clearance (mL/min) = [(140 - age ‘‘years’’) 9 (weight ‘‘kg’’) 9

(0.85 if women)]/(72 9 serum creatinine ‘‘mg/dL’’)

MDRD-4 GFR (mL/min/1.73 m2) = 186 9 [serum creatinine (mg/dL)]-1.154 9

(age ‘‘years’’)-0.203 9 (0.42 if women) 9 (1.212 if African American)

Int J Clin Pharm

123

The pharmacy department has 5 staff pharmacists who

provide drug distribution services, preparation and distri-

bution of sterile and non-sterile products, outpatient ser-

vices (including outpatient counselling), and clinical

services (review of all medication orders for interactions,

allergies, duplications, and appropriate doses). Pharmacists

also participate in several hospital committees. Pharmacists

provide these services 12 h per day, 6 days per week.

All orders are reviewed for interactive validation and

approved by ward pharmacists through the EHR and the

pharmacy system.

Population and inclusion criteria

All patients older than 18 years of age who are admitted to

medical and surgical services (excluding the intensive care

unit and the anaesthesiology ward) with an estimated glo-

merular filtration rate \60 mL/min using the MDRD4

formula were included in the study. We excluded intensive

care unit and anaesthesiology ward patients because of

different and incompatible CPOE systems. Haemodialysis

patients were also excluded.

Patients could be enrolled more than once if a new

treatment was started during the study. The patients were

consecutively included in the study.

The study was open; every patient admitted to the

medical and surgical services (excluding the intensive care

unit and the anaesthesiology ward) with an estimated GFR

\60 mL/min was included.

To identify patients with a GFR\60 mL/min, the labo-

ratory results from the EHR were integrated with the

pharmacy computer system, as were orders from the CPOE.

When measured serum creatinine was available, the

laboratory service calculated the GFR with the MDRD4

formula for each patient.

We consider a patient to be in steady-state regarding

renal function when the value of the GFR was consistent

with previous GFR values or when two consecutive GFR

values in 72 h were similar. We also excluded any possible

situation that can change creatinine serum level

temporarily.

Intervention

Using the Summary of Product Characteristics and two

drug databases (Micromedex 2.0� and Lexicomp�) spe-

cific-dosage guidelines for drugs used in patients with renal

impairment were established and entered into the pharmacy

computer system.

These guidelines allow the pharmacist to identify which

drugs require monitoring in patients with renal insuffi-

ciency and make dosage recommendations to the pre-

scribing physician.

Before the daily electronic order validation by the

pharmacist (6 days a week, Monday to Saturday), a list of

patients with a GFR \60 mL/min and a drug prescribed

that may require renal dose adjustment was displayed using

the pharmacy computer system.

To avoid over alerting, not all dosage modification alerts

were forwarded to the physician, and pharmacists reviewed

the electronic charts of the patients marked in the list and

recommend (if necessary) appropriate dosage-modifica-

tions to the prescribing physician through a specific note in

the EHR (Fig. 1). With this review, the pharmacist also

looked for any condition that could invalidate or change the

recommendation (i.e., patient on renal replacement, a

transitory renal failure, etc.).

The pharmacists used the progress notes function in the

EHR to give information on how to prescribe drugs in

renally impaired patients to the physicians. The pharmacist

reviewed these alerts 24 and 48 h later, and noted any

changes in the study database. A time point of 48 h was

chosen because it is the validation period of the comput-

erized medication order, and the Pharmacy Service is

closed on Sundays.

A semi-automated procedure was chosen for two rea-

sons. First, we wanted to improve the quality of the CDSS

in a cheap and easy way (our CDSS is provided by an

external enterprise). Second, we wanted to avoid over

alerting.

Three pharmacists participated in this program.

Study design and outcomes

We performed a quasi-experimental, one-group, pretest–

posttest study from February to May 2011.

Our primary outcome was: evaluate the efficacy of this

computer-based intervention program on the frequency of

appropriate orders.

We measured the percentage of appropriate orders

before the recommendation (pre-test) and after the rec-

ommendation (post-test).

Our secondary outcomes were:

• The frequency of appropriate and inappropriate orders

among orders that required a dosage adjustment.

• The frequency of accepted recommendations: A drug

dosage recommendation was made for each inappro-

priate order detected by the semi-automated system.

The recommendations were reported to the prescriber

through specific notes in the EHR.

Accepted recommendations were defined as changes in

dose, frequency, or drug previously recommended \48 h

ago, and were included in the study database.

Partial agreements were considered as accepted recom-

mendations. ‘Partial agreement’ was defined as a different

Int J Clin Pharm

123

dose adjustment according to any other published guideline

for dosing adjustments based on renal function.

For example, piperacillin-tazobactam’s Spanish package

insert recommend a different dose adjustment (the dose rec-

ommended in patients with a creatinine clearance of

20–40 mL/min is 4/0.5 g every 8 h) than the Micromedex 2.0�

database, (for all indications except nosocomial pneumonia and

ClCr of 20–40 mL/min, the dose recommended is 2/0.25 g

every 6 h, for nosocomial pneumonia is 3/0.75 g every 6 h).

• Description of alert type: Each recommendation was

categorized as a dose, frequency, dose and frequency,

drug change or discontinuation, and therapeutic drug

monitoring alert.

• Drugs most frequently involved in recommendations.

• Medical wards most frequently involved in

recommendations.

Statistical analysis

To show a difference in the frequency of appropriate orders

of 10 %, with alpha of 0.05 and beta of 0.1 (power of

90 %), we calculated a sample of 211 cases (drug orders)

assuming a frequency of appropriate orders of 50 %.

Because the frequency of inappropriate orders reported

in other hospital-based studies ranges from 67 % [15] to

15 % [16], we assumed a frequency of 50 %, which gives

the largest sample size.

Results are expressed as mean values with a 95 %

confidence interval, proportions, or median, as appropriate.

We assessed the effectiveness of the intervention pro-

gram by comparing the differences between the frequen-

cies of appropriate orders pre- and post-test using the

McNemar test. Statistical significance was established at

p \ 0.05. All tests were performed with the SPSS v.15

software for Windows (SPSS Inc., Chicago, IL, USA).

The McNemar test was chosen to assess whether there is

a pre-post difference within a single group, and because

responses were correlated within each patient from the pre-

test to the post-test.

We developed a secondary analysis without the cases

in which the patient was discharged before the acceptance

of the recommendation could be evaluated (early

discharges).

Fig. 1 Note in the EHR (Explanations/translations). Note: ‘‘Phar-

macy Service: Estimated glomerular filtration rate by MDRD4 of

18.11 mL/min/1.73 m2 (Cr = 2.7 mg/dL). For creatinine clearance

\20 mL/min and nosocomial pneumonia, a dose adjustment of

piperaciline/tazobactam to 2/0.25 g every 6 h is recommended. We

advised that if the renal function improves, reviewing pre-adjusted

dose drugs is necessary. Thank you. For further information, please

contact Pharmacy Service: 418470’’

Int J Clin Pharm

123

Results

A total of 171 patients were included in the study (42.7 %

male). The mean age was 76.4 years (95 %CI 74.4–78.5).

The mean value of serum creatinine concentration was

2.2 mg/dL (95 % CI 2.1–2.4).

The mean value of GFR estimated by MDRD4 was

31.3 mL/min/1.73 m2 (95 % CI 29.4–33.1).

A total of 286 drug orders were reviewed in 171

patients. Of those, 35 % prompted a dosage adjustment

recommendation in 44 % of patients. Physicians accepted

60 % of these recommendations.

The frequency of appropriate prescribing based on renal

function was 65 % before the intervention. After the inter-

vention, this frequency was 86 % (p \ 0.001) (Table 2).

The distribution of alert type is summarized in Fig. 2.

Drugs most frequently involved were levofloxacin

(19 % of the interventions), dexketoprofen (13 %), and

metformin (9 %) (Table 3).

The interventions were more frequent in the emergency

department (45 %), followed by internal medicine (22 %),

and traumatology (8 %) (Table 4).

In the secondary analysis, we excluded the cases in

which the patient was discharged before being able to

evaluate the acceptance of the recommendation (15 cases,

5.2 %). In this analysis, the frequencies of appropriate

orders before and after intervention were 68.6 vs. 90.77 %

(p \ 0.001). A total of 70.6 % of the interventions were

accepted.

The program required 30–45 min of pharmacist time per

day. The average number of patients reviewed (listed in the

automatic record) daily was 28.

Discussion

The frequency of appropriate prescribing detected prior to

intervention was 65 %, and after the intervention was 86 %.

This represents a 60 % increase over the pre-test frequency

of appropriate orders in patients with renal impairment.

Reported inappropriate frequencies in other hospital-based

studies ranged from 15, 19.9, 22.5, and 23, to 67 % [11–13,

15, 16]. These great differences can be explained by differ-

ences in guideline compliance among institutions, inconsis-

tent definitions of CKD, the type of drugs evaluated, and

differences between dosage guidelines used. In any case, they

point to a need to evaluate the situation in each institution.

In our case, the frequency of appropriate prescribing was

65 % before the intervention. This corresponds to 35 % of

inappropriate prescribing.

It can be considered consistent with that identified in

previous studies and demonstrates the need for interven-

tions that improve the renal dosing of medications.

We found several kinds of intervention programs stud-

ied in the available literature from manual reviews to a

completely-automated alert system with varying results.

We chose a mixed system that combines an automated alert

system with manual review of patient charts to avoid

unnecessary alerts and allow enhancement of the process.

According to the results, this program improved renal

dosing guideline compliance in our institution.

The frequency of accepted recommendations found was

60 % in the primary analysis and 70.6 % in the secondary

analysis (without early discharges).

This frequency in other hospital-based studies was

41.8 % [18], 52 % [13], 87.7 % [17], and 88 % [12]. Our

results are similar to those found by Montanes-Pauls et al.

(63.5 %) and Nightingale et al. (65 %) however, their

populations were significantly different. Montanes-Pauls

et al. [26] developed their study in 3 long-term care

facilities, and the creatinine clearance threshold selected

was 30 mL/min. The Nightingale et al. [21] study was

performed on a specific renal service, so the patient and

practitioner profiles were very different.

Fig. 2 Type of alert. Examples: Dose change: Initial dose of

piperaciline/tazobactam: 4/0.5 g every 6 h. For creatinine clearance

\20 mL/min and nosocomial pneumonia, a dose adjustment of

piperaciline/tazobactam to 2/0.25 g every 6 h is recommended.

Dosage interval change: Initial interval of meropenem 1 g: every

8 h. For creatinine clearance between 26 and 50 mL/min, a dosage

interval change to every 12 h is recommended. Dose and frequency

change: Initial dose of amoxicillin/clavulanic acid: 1 g/200 mg every

8 h. For creatinine clearance between 10 and 30 mL/min, a dose and

frequency change to 500/50 mg every 12 h is recommended.

Therapeutic drug-monitoring: in theses cases, the results delivered

by the hospital laboratory confirmed the wrong dose

Table 2 Contigency table for the primary result number of appro-

priate orders before and after the recommendation) (McNemar test)

Appropriate orders

before the

recommendation

(pre-test)

Appropriate orders after the

recommendation (post-test)

Row

total

Yes No

Yes 186 0 186

No 60 40 100

Colum total 246 40 286

Int J Clin Pharm

123

Our study results may be attributed to several factors.

First, the way we communicated the recommendations

was written alerts and not oral. In our institution the pro-

gress notes can be read by any health care worker, not only

physicians.

Some authors have pointed out that making recom-

mendations part of the permanent medical record com-

promises confidentiality between the pharmacist and the

physician and can be a source of alert noncompliance [17].

Second, although a pharmacist reviewed the electronic

charts to avoid unnecessary or incorrect recommendations,

physicians could reject them because of a compelling

indication for the medication despite the risk.

And third, in some cases there was an accumulation of

notes and alerts and physicians could have unintentionally

rejected the recommendations.

As in other studies, the J group, according to ATC

classification, was the most frequently involved in recom-

mendations (56 %) (Fig. 3). This result is similar to other

studies such as Peterson et al. (66 %), Sellier et al.

(54.5 %), and Alvarez Arroyo et al. (60 %) [12, 13, 18].

The emergency department was the most frequently

involved (45 %), followed by internal medicine, (22 %).

These results reflect the importance of the intervention

since the emergency department is the first contact patients

have with the institution.

Limitations

Our study has several limitations. The intervention was

conducted in only one hospital, but with CPOE and phar-

macy software fairly widespread in our region.

eGFR was calculated with the MDRD4 formula, which

is imprecise for some cases (elderly patients, for example),

and does not accurately reflect renal function in non steady-

state conditions.

Also, the CG equation is the equation most commonly

used in consensus-based medication dosing guidelines for

patients with CKD, and some studies have reported dif-

ferences in drug dosages between the CG and the MDRD

equation [27]. On other hand, another study found a high

concordance frequency in drug dosage between these

Table 3 Number of interventions by drug

Number of interventions

1 2 4 Between 8 and 16

Drug Voriconazole Tramadol hydrochloride Imipenem monohydrate/cilastatin sodium Piperacillin/tazobactam

Tranexamic acid Ciprofloxacin Hydrochlorothiazide Meropenem

Gemfibrozil Ceftazidime Enoxaparin sodium Amoxicillin and clavulanic acid

Gabapentin Cefepime Digoxin Metformin

Clarithromycin Abacavir/Lamivudine Colchicin Dexketoprofen trometamol

Cefotaxime Levofloxacin

Cefazolin

Atenolol

Table 4 Recommendations by

medical wardNumber of recommendations

\4 8 22 45

Medical Ward Cardiology Traumatology Internal

Medicine

Emergency

Department

Hematology

General and

Digestive Surgery

Oncology

Urology

Gastroenterology

Otorhinolaryngology

Nephrology

Pneumonology

Neurology

Int J Clin Pharm

123

equations [28]. Currently, there is not a consensus on

whether or not to use the MDRD equation to determine

drug dosages, although the NKDEP (National Kidney

Disease Educational Program) recommends both equations

to estimate renal function for dose adjustment [29].

We performed a quasi-experimental study design (pre-

post intervention study) because it was not logistically

feasible to conduct a randomized controlled trial, which

reduces spurious causality and bias to reach strong

conclusions.

We performed an open study because the hospital

doesn’t allow keeping any doctor out of an intervention. It

is a small hospital with high rotation of doctors between

wards.

The pharmacists involved in the interventions were

involved in the data analysis, so this might introduce bias

to the study.

We selected a primary outcome based on the analysis of

number of orders instead of patients because, in our

experience, physicians do not give the same importance to

all drugs in renal insufficiency.

In patients with several drugs prescribed, some drugs are

adjusted and others are not. Number of orders is not the

best data set, but we wanted to assess the prescribing pat-

terns for renal impairment in our institution.

We do not know if some of the alerts were uninten-

tionally rejected. It is possible that some of the recom-

mendations were rejected without a clinical cause.

We assessed the impact of the program as the frequency

of accepted recommendations rather than patient-outcome

measures, but this is an important area for future study.

To our knowledge, only two studies assessed clinical

outcomes. They demonstrated the program’s efficacy with

different clinical outcomes [16, 25].

Chertow et al. found significantly shorter length of stay

(adjusted for age, sex, and diagnosis-related group weight)

during the interventions period. In the Rind et al. study, the

intervention resulted in a decrease of relative risk of serious

renal impairment, although they excluded patients with

pre-existing moderate or severe renal insufficiency.

Another important issue is the efficiency of this kind of

program. As far as we know, only two studies have eval-

uated the cost of these programs, and they found con-

tradicting results [12, 16].

Further research is needed to elucidate these points.

Conclusion

This study found that a computer-based, semi-automated

drug-dosage program for renal-failure patients was able to

reduce the number of inappropriate orders due to renal

insufficiency.

In a considerable number of patients, the presence of

renal dysfunction is not considered in medication pre-

scribing, and that this kind of intervention may result in a

substantial reduction of inappropriate drug orders, and

consequently, may prevent adverse drug reactions.

Acknowledgments We thank Federico Tutau PharmD PhD, for his

help in the initial design of the system. We also thank the hospital

informatics department for its contribution to the design and imple-

mentation of the program.

Funding This study was supported in part by the research grant

‘Salud, Prevencion y Medio Ambiente y Seguros’, from Fundacion

MAPFRE.

Conflicts of interest None.

References

1. Gracia S, Montanes R, Bover J, Cases A, Deulofeu R, Martin de

Francisco AL, et al. Documento de consenso: Recomendaciones

sobre la utilizacion de ecuaciones para la estimacion del filtrado

glomerular en adultos. Nefrologia. 2006;26(6):658–65.

2. Martinez Castelao A, Martin de Francisco A, Gorriz J, Alcazar R,

Orte L. Estrategias en salud renal: un proyecto de la Sociedad

Espanola de Nefrologıa. Nefrologia. 2009;29(3):185–92.

3. National Kidney F. K/DOQI clinical practice guidelines for

chronic kidney disease: evaluation, classification, and stratifica-

tion. Am J Kidney Dis. 2002;39(2 Suppl 1):S1–266.

4. Cockcroft DW, Gault MH. Prediction of creatinine clearance

from serum creatinine. Nephron. 1976;16(1):31–41.

5. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A

more accurate method to estimate glomerular filtration rate from

serum creatinine: a new prediction equation. Modification of Diet

in Renal Disease Study Group. Ann Intern Med. 1999;130(6):

461–70.

6. Alcazar R, Egocheaga MI, Orte L, Lobos JM, Gonzalez Parra E,

Alvarez Guisasola F, et al. Documento de consenso SEN-sem-

FYC sobre la enfermedad renal cronica. Nefrologia. 2008;28(3):

273–82.

7. Munar MY, Singh H. Drug dosing adjustments in patients with

chronic kidney disease. Am Fam Physician. 2007;75(10):1487–96.

8. Long CL, Raebel MA, Price DW, Magid DJ. Compliance with

dosing guidelines in patients with chronic kidney disease. Ann

Pharmacother. 2004;38(5):853–8.

Fig. 3 Interventions by ATC group

Int J Clin Pharm

123

9. Kuperman GJ, Bobb A, Payne TH, Avery AJ, Gandhi TK, Burns

G, et al. Medication-related clinical decision support in com-

puterized provider order entry systems: a review. J Am Med

Inform Assoc. 2007;14(1):29–40.

10. Rubio Fernandez M, Aldaz Frances R, Garcia Gomez C, Valla-

dolid Walsh A. Caracterısticas de la prescripcion electronica a-

sistida en hospitales espanoles. Farm Hosp. 2005;29(4):236–40.

11. Sellier E, Colombet I, Sabatier B, Breton G, Nies J, Zapletal E,

et al. Effect of alerts for drug dosage adjustment in inpatients with

renal insufficiency. J Am Med Inform Assoc. 2009;16(2):203–10.

12. Alvarez Arroyo L, Climent Grana E, Bosacoma Ros N, Roca

Merono S, Perdiguero Gil M, Ordovas Baines JP, et al. Evaluacion

de un programa de intervencion farmaceutica en pacientes con

medicamentos en riesgo renal. Farm Hosp. 2009;33(3):147–54.

13. Oppenheim MI, Vidal C, Velasco FT, Boyer AG, Cooper MR,

Hayes JG et al. Impact of a computerized alert during physician

order entry on medication dosing in patients with renal impair-

ment. Proc AMIA Symp. 2002:577–81.

14. Bertsche T, Fleischer M, Pfaff J, Encke J, Czock D, Haefeli WE.

Pro-active provision of drug information as a technique to address

overdosing in intensive-care patients with renal insufficiency. Eur

J Clin Pharmacol. 2009;65(8):823–9.

15. Falconnier AD, Haefeli WE, Schoenenberger RA, Surber C,

Martin-Facklam M. Drug dosage in patients with renal failure

optimized by immediate concurrent feedback. J Gen Intern Med.

2001;16(6):369–75.

16. Chertow GM, Lee J, Kuperman GJ, Burdick E, Horsky J, Seger

DL, et al. Guided medication dosing for inpatients with renal

insufficiency. JAMA. 2001;286(22):2839–44.

17. Peterson JP, Colucci VJ, Schiff SE. Using serum creatinine

concentrations to screen for inappropriate dosage of renally

eliminated drugs. Am J Hosp Pharm. 1991;48(9):1962–4.

18. Gea Rodriguez E, Barral Vinals N, Manso Mardones P, Indo

Berges O. Contribucion a la seguirdad en la utilizacion de hep-

arina de bajo peso molecular en pacientes con insuficiencia renal.

Farm Hosp. 2004;28(2):101–5.

19. Goldberg DE, Baardsgaard G, Johnson MT, Jolowsky CM, Shep-

herd M, Peterson CD. Computer-based program for identifying

medication orders requiring dosage modification based on renal

function. Am J Hosp Pharm. 1991;48(9):1965–9.

20. McMullin ST, Reichley RM, Kahn MG, Dunagan WC, Bailey

TC. Automated system for identifying potential dosage problems

at a large university hospital. Am J Health Syst Pharm.

1997;54(5):545–9.

21. Nightingale PG, Adu D, Richards NT, Peters M. Implementation

of rules based computerised bedside prescribing and adminis-

tration: intervention study. BMJ. 2000;320(7237):750–3.

22. Galanter WL, Didomenico RJ, Polikaitis A. A trial of automated

decision support alerts for contraindicated medications using

computerized physician order entry. J Am Med Inform Assoc.

2005;12(3):269–74.

23. Field TS, Rochon P, Lee M, Gavendo L, Baril JL, Gurwitz JH.

Computerized clinical decision support during medication

ordering for long-term care residents with renal insufficiency.

J Am Med Inform Assoc. 2009;16(4):480–5.

24. Matsumura Y, Yamaguchi T, Hasegawa H, Yoshihara K, Zhang

Q, Mineno T, et al. Alert system for inappropriate prescriptions

relating to patients’ clinical condition. Methods Inf Med.

2009;48(6):566–73.

25. Rind DM, Safran C, Phillips RS, Wang Q, Calkins DR, Delbanco TL,

et al. Effect of computer-based alerts on the treatment and outcomes

of hospitalized patients. Arch Intern Med. 1994;154(13):1511–7.

26. Montanes-Pauls B, Saez-Lleo C, Martinez-Romero G. Ajuste de

dosificacion de medicamentos en pacientes ancianos institucio-

nalizados con insuficiencia renal. Farm Hosp. 2009;33(1):43–7.

27. Wargo KA, Eiland EH 3rd, Hamm W, English TM, Phillippe

HM. Comparison of the modification of diet in renal disease and

Cockcroft-Gault equations for antimicrobial dosage adjustments.

Ann Pharmacother. 2006;40(7–8):1248–53.

28. Stevens LA, Nolin TD, Richardson MM, Feldman HI, Lewis JB,

Rodby R, et al. Comparison of drug dosing recommendations

based on measured GFR and kidney function estimating equa-

tions. Chronic Kidney Disease Epidemiology Collaboration. Am

J Kidney Dis. 2009;54(1):33–42.

29. The National Kidney Disease Educational Program. CKD and

drug dosing: information for providers. [Internet] 2009 Sep

[updated 2012 March; cited 2013 July 4]. Available from http://

www.nkdep.nih.gov/resources/CKD-drug-dosing.shtml.

Int J Clin Pharm

123