Ethical dilemmas in antibiotic prescribing: analysis of everyday practice

Ethical dilemmas in antibiotic prescribing: analysisof everyday practice

I. Bjornsdottir* PhD (Pharm) and E. H. Hansen MSc (Pharm)*CEO, The Pharmaceutical Society of Iceland, Holtaseli 36, IS-109 Reykjav k, Iceland and Professor,Department of Social Pharmacy, Royal Danish School of Pharmacy and Director, FKL-Research Centre forQuality in Medicine Use, Copenhagen, Denmark

SUMMARY

Objective: To explore general practitioners

(GPs) views on their obligations with respect to

diagnosing infections and prescribing antibiotics.

Methods: The GPs reflections and prioritization

were studied by means of interviews and obser-

vations. We analysed how their prioritization

complied with an ethical guidance that ranked

patient autonomy and welfare highest, then

competence obligations and obligations to soci-

ety, followed by fraternal obligations.

Results: Balancing of pros and cons was promin-

ent in our informants decision making but often

resulted in decisions that deviated from the ethical

guidance. The ranking varied much between the

GPs. The highest priorities in the GPs practice

were related to the patients everyday life (some-

times autonomy, sometimes beneficence in a broad

sense), doctorpatient relationship (communica-

tion competence), the patients perceived import-

ance on the job market (society) and relationship

with colleagues (fraternal). Perceived lack of

resources and uncertainty with respect to both

diagnostic and treatment decisions frequently

influenced decision making.

Keywords: antibiotics, general practice, Iceland,

obligations, prioritization, qualitative methods

INTRODUCTION

Physicians often experience discomfort regarding

decisions on antibiotic prescribing, and also when

internal rules are broken, even when clashes

between rules make it impossible to comply with

all of them (1).

Antibiotics have been classified as therapeutic

trial drugs, because, although they are supposed to

be prescribed only on the basis of certain diagnosis

of bacterial infections, real life circumstances often

make certainty in diagnosing difficult or imposs-

ible (25).

Ethical problems in everyday practice have not

gained much bioethical attention (6, 7). The big

issues have been prioritization, end of life decisions

and recent advances in biotechnology, although

physicians role and behaviour, gate-keeping in

health care, usability of guidelines and paternalism

vs. autonomy have been discussed (814).

It still remains uncertain whether the teleological

(main emphasis on the outcome, maximizing hap-

piness, joy or good), the deontological (main

emphasis on the means, having virtuous reasons

for doing the right thing) or one of the varieties of

other ethical theories is most applicable (15).

Although the doctorpatient relationship has

traditionally been structured around the clinical

model, which has been described as utilitarian,

outcome or ends is frequently rather unpredictable

in health care, which can make a relational model

(deontological in essence) more appropriate in

many cases (16). In their Principlism Theory,

Beauchamp and Childress identify beneficence,

non-maleficence, autonomy and or justice as thefundamentals of health care ethics (17). Physicians

have been found to act in accordance with benefi-

cence principles rather through recognition of

expertise than because of presumed patient

vulnerability and lack of understanding, and they

Received 29 August 2002, Accepted 10 October 2002

Correspondence: Ingunn Bjornsdottir PhD (Pharm.), CEO,

The Pharmaceutical Society of Iceland, Box 252, IS-172 Seltjar-

narnes, Iceland. Tel: +354 561 6166; fax: +354 561 6182; e-mail:

ingunnbj@itn.is

An earlier version of this paper was a part of the results section

of a PhD thesis by the first author. It was defended in August

1999, at the Royal Danish School of Pharmacy, Department of

Social Pharmacy.

Journal of Clinical Pharmacy and Therapeutics (2002) 27, 431440

2002 Blackwell Science Ltd 431

have been found to be inconsistent in their attitude

towards patient autonomy (18, 19). Furthermore,

when describing experienced ethical dilemmas,

they seem to use a somewhat broader definition

than the bioethics literature (added concerns about

own reputation and doctorpatient relationship to

the mainstream definition of conflict and choice

between values, beliefs and options for action) (20).

We have described Icelandic general practitioners

(GPs) rationales for prescribing antibiotics earlier

(21). We found that their primary purpose of pre-

scribing antibiotics was to help patients to carry on

with their everyday activities, i.e. outcome-oriented

or teleological thinking. Lack of resources, or other

reasons, however, could cause the GPs to deviate

from evidence-based medicine. The prescribing

could occasionally be a result of respecting the

autonomy of the patient at the expense of benefi-

cence or non-maleficence, i.e. process-oriented,

deontological thinking, because of lack of time or

access to technology, or a belief in the patients

right to decide (21). Codes of ethics in medicine fail

to deal with the question of resources (22).

Moral obligations of health care professionals

can be divided into four groups: primary

obligations (respect for autonomy, consideration of

beneficence), preservation maintenance of profes-sional competence (necessary for the ability to

prioritize when autonomy, beneficence, non-

maleficence and or justice clash), societal (publichealth, justice) and or fraternal (23).

OBJECTIVES

To explore

GPs views on their obligations with respect toinfections and antibiotics.

How GPs prioritize when clashes between per-ceived obligations occur.

MATERIALS AND METHODS

The study was conducted in Iceland, a relatively

sparsely populated country, with its roughly

280 000 inhabitants, approximately 70% of whom

live in the capital and surroundings, where the

largest hospitals and most of the specialists are also

situated. The remaining 30% of the inhabitants live

mostly in the coastal areas (rural) where GPs are

the key actors in the health service. Some solo GPs

cover an area with travel distances of up to 100 km.

Physicians have mainly specialized in general

practice in Sweden. Female GPs have been relat-

ively few, as have contractors.

Qualitative methods (in-depth interviews and

observations) were used to grasp the content of the

GPs considerations and reflections.

Interviews

The sample (10 GPs) was purposefully selected to

reflect existing variations in the GPs age and years

of professional experience (two 55 years, mean age 48), gender (two females),

practice organization (eight state-employed, two

contractors), practice size (two in solo practice,

eight in group practice), practice location (three

rural, seven urban) and postgraduate training (six

specialized in Sweden, one in Canada, three did

not specialize), as these factors might influence

prescribing habits (24).

Informants were added to the sample until

the data were saturated. Informants and patients

exposed to observation, gave informed ver-

bal consent. The study had ethical committee

approval.

The interview-guide was aimed at exploring the

doctors perceived reasons for antibiotic prescri-

bing. After the first observation, it focused on atti-

tudes, common infections (symptoms diagnosisand treatment), patient variables (e.g. age and

gender), recent antibiotic prescriptions and poss-

ible associated discomfort, pressure from patients,

resources and co-operation with other health care

workers. The physicians were encouraged to vol-

unteer any information they found important.

Specialists was an issue initiated by the

observed interviewed physician in more than oneof the first interviews and hence added to the

interview-guide. The interviews lasted for 45 min

to 2 h, and were tape-recorded and transcribed in

full. One informant did not allow tape recording,

but detailed notes were kept instead.

Observations

Three of the GPs were observed for 310 h each

(rural, urban, solo practice and two different size

group practices), each one before he was inter-

viewed, creating a basis for the interview guide

432 I. Bjornsdottir and E. H. Hansen

2002 Blackwell Science Ltd, Journal of Clinical Pharmacy and Therapeutics, 27, 431440

and subsequent modification of it. Detailed field

notes were taken and typed shortly after the

observation. A total of about 60 doctorpatient

contacts were observed (including telephone con-

tacts). Antibiotics and or infections were discussedin 15 contacts. Three antibiotic prescriptions were

issued.

Our data seemed saturated (25) after approxi-

mately seven interviews (i.e. interviews eight to 10

did not seem to add themes or nuances). This early

saturation might be due to the use of the observa-

tions in creating the interview guide and, perhaps,

also because of the data collectors (first authors)

experience from primary health care (10 years in

community pharmacies, with daily contact with

GPs, with whom antibiotic prescribing was

frequently discussed).

Data on ethical dilemmas were extracted and

analysed by open, axial and selective coding

(grounded theory procedures), but aiming at an

in-depth understanding of the informants rather

than developing concepts and theories (25, 26).

The interviewers possible biases were minimized

by researcher triangulation (authors, input from

two sociologists in the open coding). The

informants offered information about incidents

they could have handled better (such incidents

were also observed). Hawthorne effects were

therefore considered minimal (27). The results

are the researchers interpretation of the physicians

attitudes and behaviour (28), but as a validity

check, the results were shown to two physicians

with experience from general practice and pre-

sented at conferences where GPs were among

the audience. These physicians confirmed the

analysis.

RESULTS

Balancing of scientific and practical considerations

was prominent in the GPs decision making. Many

of them described individual guidelines, or rules

of a thumb that they had developed to work out

their practices. They considered this development

of own style acceptable, even preferable. One

doctor described this as bringing the art into the

work. Perceived pressure from patients, society

(e.g. lack of resources) and or other health careprofessionals was interwoven into practically all

their decision making.

The inspiration for the order of our presentation of

the GPs prioritization is the classification of pro-

fessional obligations, described in the Introduction.

Primary obligations towards the client

The GPs tendency to refer to their clients as peo-

ple rather than patients (emphasizing the person

rather than the case) was reflected in their decision

making.

Respect for autonomy. The GPs generally paid

regard to patient preferences when deciding on

diagnostic procedures or treatment, but varying

from respecting her right to refusal to meeting her

demands. Co-operation with patients was consid-

ered important, especially regarding treatment.

The use of the wait-and-see method depended on

the patients preferences.

An easy access to the GP (appointment within a

few hours) was considered an important patient

right, although not necessarily justified by health-

related needs:

when you let someone diagnose there is of course a

risk of a wrong diagnosis, but letting you see itdictate your getting to the doctor, whether you get

wrong treatment or not, of course that is no good,

you see. One has a right to get to the doctor when

the need arises A

It sometimes seemed unclear, whether respect for

the patients autonomy influenced the GPs deci-

sions. For example, when they helped patients to

avoid sick leave, they sometimes did so on the

basis of beneficence rather than patient request.

Beneficence non-maleficence. The GPs defined bene-ficence non-maleficence in a broad sense, not justin terms of health. Their arguments for different

service levels for different patient groups were

circumstances in the patients life in general, rather

than specifically their health.

1. Job: The patients vulnerability on the job market

was considered during decision making. Children

were considered to need quick service, because of

parents limited rights (generally 7 days year) tostay home with ill children. People, who were

regarded to be at risk of losing their jobs because of

sick leaves, were also considered to need quick

service.

2002 Blackwell Science Ltd, Journal of Clinical Pharmacy and Therapeutics, 27, 431440

Ethical dilemmas in antibiotic prescribing 433

2. Nuisance: Doctors in the rural area thought they

ordered fewer to-be-certain investigations than

colleagues in the large towns, to avoid unnecessary

hassle for the patient (implying that avoiding has-

sle might be more important than certainty in

diagnosing). A doctor working in a rural area felt

that he had to keep admittance to a hospital low,

because of the inconvenience inherent in a hospital

stay, far away from home.

3. Economy: Some GPs stated that they would

omit a test if (in their opinion) the patient could not

really afford it.

Actually, the price is only 900 kronur (approxi-

mately 9 US$). But [] if you have three kids in arow, and if you intend to take a swab, then you

dont take a swab from the whole row, not unless

you have tried to treat and it doesnt work D

A GP who used tests, sometimes forgot to bill

people for them, thereby ranking the unbilled

individuals economy higher than cost to society.

Professional competence

This section deals with background knowledge,

technical skills and communication skills.

Background knowledge. Own experience from work

in hospitals and discussions with older colleagues

were considered very important sources of

knowledge, seemingly more important than

formal education. Hence, varying experience is

among the causes of variations in decision

making.

Technical skills. Technical skills were perceived to

increase with increasing experience, justifying

highly individual necessity for evidence, to confirm

a diagnosis.

I think this is a very personal style [] what weare taught, really, is that for sinusitis then this

treatment is needed and so on, sinusitis can be

diagnosed by this by X-ray and then perhapswhen one gets more confident in the clinic, then

one can perhaps allow oneself a bit more C

One doctor described how he had learned to use

the smell and appearance of a urine sample to

distinguish between bacteria. He did not elaborate

on the success of that method as compared with

tests.

Educated guesses regarding causative agents in

adults occurred. For example, urinary tract infec-

tion (UTI) causing bacteria in elderly women were

considered likely to be different from those causing

UTI in younger women, based on the womens

assumed (lack of) sexual activity. Examples were

also given based on the educated guess on the

patients job.

We are a bit in connection with the basic industries

here [] this is just a feeling really, but I get manyfishermen who get hooks i...