Nigerian Journal of Paediatrics

A survey of paediatricians on evidence-based medical practice in Nigeria

Niger J Paediatr 2018; 45 (3): 163 – 170

ORIGINAL

Ogunlesi TA

CC – BY A survey of Paediatricians on

evidence-based medical practice in

Nigeria

DOI:http://dx.doi.org/10.4314/njp.v45i3.5

Accepted: 17th September 2018

Abstract : Background: The in-

clinical practice while 53.9% be-

creasing need for a uniform stan-

lieved that EBM is more suitable

Ogunlesi TA (

)

dard of medical practice necessi-

for the developed world. The over-

Department of Paediatrics,

tates an examination of the im-

all mean knowledge score was

Olabisi Onabanjo University

pediments to the use of evidence-

77.6%. Knowledge about EBM

Teaching Hospital,

based medicine.

was rated high among 80% of the

P. O. Box 652, Sagamu-121001NG,

Objective: To examine the knowl-

participants and the proportion

Ogun State, Nigeria

edge, perceived barriers to the use

with high knowledge was not sig-

Email:

of evidence-based medicine and

nificantly associated with EBM

tinuade_ogunlesi@yahoo.co.uk

the associated profession-related

use in routine practice or prior

factors among paediatricians in

EBM training. Some of the per-

Nigeria.

ceived barriers to EBM use in-

Methods: A cross-sectional sur-

cluded lack of skills for under-

vey of paediatricians in Nigeria

standing statistics (81.6%), lack of

was conducted between March

incentives for the use of EBM

and April 2013 by Email. The

(64.7%), lack of time (44.6%) and

knowledge about evidence-based

lack of conclusive evidences

medicine and systematic reviews

(44.6%).

was tested using a structured

Conclusion: The knowledge of

questionnaire and the responses

EBM among Nigerian paediatri-

were provided on three-item

cians was good but a high propor-

Likert scale.

tion regarded lack of skills to per-

Results: The response rate in the

form statistical analysis required

survey was 56.5% (65/115). Al-

for EBM as a major barrier to the

most 37% of the participants had

practice of EBM.

postgraduate qualification in or

before year 1999 (Group I). Ma-

Key words: Evidence-based medi-

jority (93.8%) of the respondents

cine, Systematic review, Research,

believed that EBM should form

Paediatricians.

the basis for decision making in

Introduction

access to a large volume of pockets of research outputs

from various parts of the world, some with conflicting

The prime role of research in driving clinical practice

reports and some with inadequate information to guide

practice.

2,3

has not changed over the years but the application of

Therefore, evidence-based medicine appears

research in guiding clinical practice is globally getting

to be the most efficient way of delivering clinical care in

focused on the critical need for evidences guiding every

the context of rapidly changing trends in medical prac-

aspect of decision taking in medical practice. The con-

tice.

tents of textbooks, clinical observations and opinions

passed down from senior colleagues have hitherto,

At the tertiary level of care, the tripod of clinical duties

formed the bedrock of guidance for making clinical de-

includes service, teaching and research. A previous

study conducted among medical specialists in Nigeria

cisions. With the increasing use of evidence-based

medicine, clinical decisions are gradually shifting to-

showed that research function was ranked third by

wards the provisions of management protocols and treat-

64.7% of the participants in that study, behind service

ment guidelines rather than textbook recommendations

and training. A similar study revealed that the spectrum

and opinions formed in the course of training and prac-

of research tilted more towards low-budget cross-

tice. In brief, evidence-based medicine strictly entails

sectional surveys and retrospective studies due to the

adherence to the current best practices for the utmost

challenges of lack of funds, work overload, time con-

straints and insufficient infrastructural supports. In the

benefit of the patient. This is necessitated by the need

for a more guided approach rather than the haphazard

same study, 37.5% of specialists had never been in-

164

volved in clinical trials. Incidentally, randomized clini-

resulting in 115 rather than the targeted 104. The se-

cal trials, which represent the peak of clinical evidence

lected 115 paediatricians were enrolled into survey and

through the benefits of elimination of bias, form the core

were contacted via email. The email contained a state-

of evidence-based medicine.

ment of introduction of the study as well as the request

Evidence-based medicine entails the identification of an

for consent. The email message specified that response

intervention or relationship of interest, gathering rele-

to the survey meant consent for enrolment into the study

vant high-quality randomized controlled trials and con-

and anyone not willing to participate in the study was

ducting statistical analysis, using meta-analysis, to as-

allowed to disregard the request. The data collection was

carried out between 3 March and 5 April 2013.

sess the quality of evidence in support of specific out-

come variables of interest.

1,6

This is the basis of the use

The research tool was a self-designed close ended ques-

of various treatment guidelines and management proto-

tionnaire with three sections using information gathered

cols for clinical decision in various parts of the devel-

from the Cochrane Handbook for Systematic Reviews of

Interventions as a template in most cases.

oped world. Commonly accessed resources for evidence

The first

-based medicine include the Cochrane Library and the

section obtained data on professional parameters such as

World Health Organisation Reproductive Health Li-

the year of postgraduate qualification, location of prac-

brary. These are the commonly used databases which

tice in terms of geopolitical zone, sector of practice

provide information on virtually all aspects of medicine.

(public or private), setting of practice (academic or non-

Of utmost relevance to the practice of paediatrics are the

academic), current use of EBM in clinical practice and

Neonatal and the Infectious Diseases Review Groups

history of previous formal training in EBM. The second

which presently contains hundreds of systematic re-

section tested the knowledge of participants about EBM

views.

using general statements covering a wide scope of the

principles and tenets of EBM as well as test of under-

With the drive to develop treatment protocols for com-

standing and ability to teach specific statistical items.

mon paediatric disorders in Nigeria, it is essential to

The third section assessed the participants’ attitude to

examine the acceptability and understanding of the con-

perceived barriers to the use of EBM in clinical practice.

cept of evidence-based medicine among paediatricians.

Some of the statements were framed in the positive con-

Therefore, the objective of this study was to examine the

text and others were framed in the negative context and

knowledge, perceived barriers to the use of evidence-

the responses were obtained using a three-item Likert

based medicine and the associated profession-related

Scale – Agreed, Undecided and Disagreed. The re-

factors among paediatricians in Nigeria.

sponses to each statement were uniformly assessed with-

out weighting; they were scored 3, 2 and 1 depending on

the context of positivity or negativity; the option of

“agreed” for a negatively -framed statement earned a

Materials and Methods

score of “1” just as a response of “disagreed” for a nega-

tively- framed statement earned a score of “3” and vice

This was a cross-sectional questionnaire-based survey of

versa . For each participant, the total score was con-

paediatricians who practiced in Nigeria. The study was

verted to percentage based on the number of statements

carried out electronically using email-distribution of

responded to. Knowledge scores of 75% and above were

questionnaire to paediatricians on a mailing list. The

classified as “high” while scores less than 75% were

research was conducted in agreement with the Helsinki

classified as “low” scores. The mean percentage knowl-

Declaration for Human subjects research. The inclusion

edge score was determined for each group of partici-

criteria included Postgraduate Fellowship in Paediatrics

pants.

obtained from either the West African College of Physi-

cians or the National Postgraduate Medical College of

Data management

Nigeria and employment as a Consultant Paediatrician

within Nigeria. The study was carried out between

Only completed questionnaires returned by email were

March and April 2013.

pooled for analysis using a spreadsheet created with the

Microsoft Excel software. Descriptive and inferential

The minimum sample size was determined using the

statistics were conducted using the SPSS version 20.0

formula; n = (z ×p×q)/d where z = 1.96, p = 0.05

software. Hypotheses were tested using the Chi Square

(proportion of respondents in a similar study who used

test with either Yate’s correction or the Fisher’s Exact

EBM to take clinical decisions ), q = 0.95 and d = 0.05.

test as necessary, for proportions of categorical variables

The calculated minimum sample size was 72 but addi-

and the Student’s t -test for the means (±Standard devia-

tional 50% (36) was added to increase the strength of the

tions) of continuous variables. Professional characteris-

study. Therefore, the final sample size was 104.

tics of the respondents (year of postgraduate qualifica-

Out of the 426 names on the mailing list of paediatri-

tion, sector of practice, setting of practice, use of EBM

cians, 78 duplicated names, names with incomplete data

in clinical practice and training in EBM) were related to

(such as place of practice and year of postgraduate quali-

the knowledge score as well as the perceived barriers to

fication) and names of deceased paediatricians were

EBM. P values less than 0.05 were accepted as statisti-

removed, leaving 348 names. Thereafter, the 348 names

cal significance.

were arranged alphabetically and systematic random

sampling was done using a sampling interval of three

165

Results

world (53.9%).

General description of the respondents

Table 1: Knowledge and attitude of respondents to EBM

Sixty-five out of the 115 participants returned com-

Suggestions about EBM

Agreed

Unde-

Dis-

Total

pletely filled questionnaire giving a response rate of

cided

agreed

56.5%. The 65 paediatricians were distributed across the

Current research findings

0 (0.0)

geo-political zones of the country as follows: 24

are useful in the day-to-day

(100.0)

management of my patients

(36.9%) from the south-west, 16 (24.6%) from the south

Adoption of evidence-based

4 (6.2)

-east, 10 (15.5%) from the south-south, 6 (9.2%) each

practice places too many

(35.4)

(56.4)

from the north-west and north-central and 3 (4.6%) from

demands on my workload

the north-east.

EBM is patient centered

4 (6.2)

5 (7.7)

(86.1)

EBM is of limited value in

3 (4.6)

4 (6.2)

The distribution of the respondents according to the year

paediatric practice

(89.3)

of post-graduate qualification was as follows: 4 (6.1%),

Literature and research

3 (4.6)

0 (0.0)

20 (30.8%), 30 (46.2%) and 11 (16.9%) for 1980-1989,

findings are useful in paedi-

(95.4)

atric practice

1990-1999, 2000-2009 and >2010 respectively. They

I need to increase the use of

15 (23.1)

0 (0.0)

were re-grouped into two as 1980-1999 (Group I) and

evidence in my daily prac-

(76.9)

2000 and above (Group II) translating to 24 (36.9%) in

tice

Group I and 41 (63.1%) in Group II. The major sub-

I am interested in learning

5 (7.7)

0 (0.0)

skills to incorporate EBM in

(92.3)

specialties included neonatology (17; 26.2%), cardiol-

my practice

ogy (9,13.9%), haematology and nephrology (7; 10.8%

EBM improves the quality

6 (9.2)

1 (1.5)

each), endocrinology, infectious diseases and ambula-

of patient care

(89.3)

tory paediatrics (6; 9.2% each). The remaining 7

EBM does not take into

7 (10.8)

consideration the limitations

(55.4)

(33.8)

(10.7%) included neurodevelopmental paediatrics, gas-

in clinical practice

troenterology and nutrition and respiratology.

Reimbursement rate will

40 (61.4)

increase with incorporation

(18.5)

(19.9)

Most of the respondents (53; 81.5%) practiced in the

of EBM into paediatric

practice

public sector; 33 (50.8%) and 29 (44.6%) practiced in

Strong evidence is lacking

15 (23.0)

Teaching Hospital and specialist hospitals (including

in most interventions used

(35.5)

(41.5)

Federal Medical Centres) respectively. The settings of

EBM should form the basis

2 (3.1)

practice were sub-classified into two: academic (33;

of decision-making in clini-

(93.8)

cal practice

50.8%) and non-academic (32; 49.2%). Fifty-two

EBM does not take into

7 (10.8)

(80.0%) were involved in undergraduate training while

consideration, patients’

(38.5)

(50.7)

all the respondents had experience with postgraduate

preferences

training. Fifty-nine (90.8%) respondents used evidence-

EBM is useful in designing

11 (16.9)

0 (0.0)

the medical curriculum

(83.1)

based medicine in their routine clinical practice but only

EBM is most useful for

11 (16.9)

24 (36.9%) actually had training in evidence-based

paediatric practice in the

(53.9)

(29.2)

medicine. While all the respondents in Group I (n = 24)

developed world

used EBM in clinical practice, 35 (85.4%) of Group II

used EBM in clinical practice (Fisher’s Exact Test =

EBM – Evidence-based Medicine

5.872; p =0.07). A significantly larger proportion of

respondents in Group I (17; 70.8%) received training on

In the assessment of the skills required to use EBM,

EBM compared to respondents sin Group II (7; 17.1%)

Table 2 shows that some of the respondents did not un-

( χ = 18.786; p < 0.001).

derstand but will like to learn heterogeneity (44.6%)

while 46.1%, 52.3%, 50.8%, 46.1%, 50.8% and 64.6%

General knowledge and attitude of respondents about

understood meta-analysis, systematic review, assess-

EBM

ment of publication bias, literature search strategies,

study designs and clinical significance of study findings

As shown in Table 1, all the respondents agreed that

respectively. The highest proportion of the respondents

current research findings are useful in routine care of

demonstrated the willingness to teach literature search

patients, 76.9% craved for further use of EBM in their

strategies (41.5%) while only 13.8% understood and

daily practices while 92.3% were willing to learn the

could teach the use of Cochrane Library Database.

skills required to incorporate EBM into their routine

practices. However, 86.1% believed EBM is patient-

In Table 3 lack of the skills for understanding statistics

centred, improves the quality of care available to pa-

and lack of skills for locating best research evidence

tients (89.3%), should form the basis of decision making

were the leading perceived barriers to the use of EBM as

in clinical practice (93.8%) and should improve the de-

identified by 81.6% and 73.8% respectively of the re-

sign of medical curriculum (83.1%). On the other hand,

spondents. The fear of medicolegal tussle arising from

the respondents identified drawbacks such as EBM not

the use of EBM was identified by only 27.7% as a bar-

taking into consideration the limitations in clinical prac-

rier to the use of EBM. The proportions of respondents

tice (55.4%) and EBM being more suitable for practice

who agreed or disagreed with lack of time and lack of

in the developed world compared to the developing

conclusive evidence as barriers to the use of EBM were

almost comparable (44.6%). All the respondents agreed

166

that EBM should be incorporated into Continuing Medi-

Table 4: Comparison of mean knowledge scores among com-

cal Education activities and into postgraduate medical

parison groups

curriculum while only 56 (86.2%) agreed that EBM

Parameters

Groups

Mean

should be incorporated into medical undergraduate train-

scores (%)

values

ing curriculum whereas 9 (13.8%) were neutral.

Duration of

Group I

79.0 ± 4.5

1.583

0.118

practice

Table 2: Assessment of the skills required for the use of EBM

Group II

76.7 ± 6.1

among the respondents

Sector of prac-

Public

78.1 ± 5.4

1.347

0.183

It is not

I do not

I have

I under-

tice

important

under-

some

stand and

Private

75.6 ± 0.4

I under-

stand but

under-

could

stand

will like

standing

teach

Setting of prac-

Academic

76.5 ± 4.5

-1.604

0.114

others

tice

Relative Risk

0 (0.0)

23 (35.4)

19 (29.3)

Non-

76.0 ± 1.4

Meta-analysis

0 (0.0)

21 (32.3)

30 (46.1)

14 (21.5)

academic

Systematic Re-

0 (0.0)

15 (23.1)

34 (52.3)

16 (24.6)

view

EBM Use in

Yes

77.8 ± 5.8

0.734

0.466

Heterogeneity

0 (0.0)

29 (44.6)

20 (30.8)

16 (24.6)

clinical practice

Publication bias

0 (0.0)

20 (30.8)

33 (50.8)

12 (18.5)

76.0 ± 1.4

Literature Search

0 (0.0)

8 (2.3)

30 (46.1)

27 (41.5)

EBM Training

Yes

78.6 ± 4.4

1.087

0.281

Strategies

Study designs

0 (0.0)

10 (15.4)

33 (50.8)

22 (33.8)

77.0 ± 6.2

Evaluating the

0 (0.0)

21 (32.3)

26 (40.0)

18 (27.7)

validity of a study

The clinical sig-

0 (0.0)

7 (10.8)

42 (64.6)

16 (24.6)

Table 5: Relationship between the professional characteristics

nificance of study

of the respondents and their knowledge of EBM

results

Character-

High Score

Low

Statistics

Using the Coch-

0 (0.0)

25 (38.5)

31 (47.7)

9 (13.8)

istics

Score

rane Library Data-

base

Duration

Group I

22 (91.7)

2 (8.3)

χ2 = 2.184;

of practice

(n = 24)

p = 0.139*

EBM – Evidence-based Medicine

Group II

30 (73.2)

11 (26.8)

(n = 41)

Table 3: Perceived barriers to the use of EBM

Sector of

Private

9 (75.0)

3 (25.0)

χ2= 0.006; p

practice

(n = 12)

= 0.936*

Agreed

Unde-

Dis-

Public

43 (81.1)

10 (18.9)

cided

agreed

(n = 53)

Lack of time

29 (44.6)

6 (9.2)

30 (46.2)

Setting of

Academic

27 (81.8)

6 (18.2)

χ2 = 0.138;

Lack of conclusive evidence

29 (44.6)

8 (12.3)

28 (43.1)

practice

(n = 33)

p = 0.710

Lack of computing resources

37 (56.9)

5 (7.7)

23 (35.4)

Non-

25 (78.1)

7 (21.9)

academic

Lack of access to electronic

37 (57.0)

8 (12.2)

20 (30.8)

(n = 32)

databases

EBM Use

No (n = 6)

6 (100.0)

0 (0.0)

FE = 1.562;

Lack of skills for locating

48 (73.8)

5 (7.7)

12 (18.5)

in practice

p = 0.335

best research evidence

Yes (n =

46 (78.0)

13 (22.0)

Lack of skills for under-

53 (81.6)

0 (0.0)

12 (18.4)

59)

standing statistics

EBM

No (n = 41)

32 (78.0)

9 (22.0)

χ2 = 0.037;

Lack of incentive for using

22 (64.7)

8 (12.3)

15 (23.0)

Training

p = 0.847*

EBM

Yes

20 (83.3)

4 (16.7)

Fear of medico-legal tussles

18 (27.7)

36 (55.4)

(n = 24)

arising from practice

(16.9)

*Yate’s Correction applied; FE =Fisher’s Exact Test

EBM – Evidence-based Medicine

Assessment of knowledge of the respondents

Relationship between perceived barriers to EBM and

professional characteristics of the respondents

The knowledge scores for the entire study population

ranged between 64.0% and 88.0% with overall mean

In Table 6a, the proportions of respondents who agreed

score of 77.6% ± 5.6%. The mean scores for the various

that lack of time was a barrier to EBM were comparable

groups of participants were comparable as shown in

across groups. Significantly higher proportions of Group

Table 4. The assessment of knowledge showed that 52

I respondents and respondents who were trained in EBM

(80.0%) and 13 (20.0%) participants had high and low

disagreed that lack of conclusive evidence was a barrier

knowledge scores respectively. Table 5 shows that the

to EBM (p = 0.025 and p = 0.004 respectively). Higher

proportions of respondents with high knowledge score

proportions of respondents who worked in non-

were comparable across groups: duration of practice (p

academic settings and those who were trained in EBM

= 0.139), sector of practice (p = 0.936) and setting of

disagreed about lack of computing resources being a

practice (p = 0.710). In addition, EBM use in practice

barrier to EBM (p = 0.045 and p = 0.023 respectively).

and prior EBM training were also not significantly asso-

Higher proportions of respondents in Group I, those in

ciated with high knowledge scores (p = 0.335 and p =

non-academic settings and those who were trained on

0.947 respectively).

EBM also disagreed that lack of access to electronic

167

database was a barrier to EBM use (p = 0.007, p = 0.002

and p < 0.001 respectively).

Table 6a: Relationship between professional characteristics of the respondents and the perceived barriers to the use of EBM in

Paediatric Practice

Lack of time

Agreed

Undecided

Disagreed

Statistics

Duration

Group I(n = 24

12 (50.0)

0 (0.0)

12 (50.0)

FE = 3.819;

Group II (n = 41)

17 (41.5)

6 (14.6)

18 (43.9)

P = 0.155

Setting

Academic (n= 33)

15 (45.5)

2 (6.0)

16 (48.5)

χ2 = 0.664;

Non-Academic (n = 32)

14 (43.8)

4 (12.5)

14 (43.8)

P = 0.766

EBM Trained

No (n = 41)

19 (46.3)

3 (7.4)

19 (46.3)

χ2 = 0.561;

Yes (n = 24)

10 (41.7)

3 (12.5)

11 (45.8)

P = 0.804

Lack of conclusive evidence

Duration

Group I(n = 24

10 (41.7)

0 (0.0)

14 (58.3)

FE = 7.041;

Group II (n = 41)

19 (46.3)

8 (19.6)

14 (34.1)

P = 0.025

Setting

Academic (n= 33)

14 (42.4)

6 (18.2)

13 (39.4)

χ2 = 2.162;

Non-Academic (n = 32)

15 (46.9)

2 (6.2)

15 (46.9)

P = 0.376

EBM Trained

No (n = 41)

21 (51.2)

8 (0.0)

12 (29.3)

FE = 10.675;

Yes (n = 24)

8 (33.3)

0 (0.0)

16 (66.7)

P = 0.004

Lack of computing resources

Duration

Group I (n = 24

12 (50.0)

0 (0.0)

12 (50.0)

FE = 5.190;

Group II (n = 41)

25 (61.0)

5 (1.2)

11 (26.8)

P = 0.06

Setting

Academic (n= 33)

19 (57.6)

5 (15.2)

9 (27.2)

FE = 5.958;

Non-Academic (n = 32)

18 (56.2)

0 (0.0)

14 (43.8)

P = 0.045

EBM Trained

No (n = 41)

26 (63.4)

5 (12.2)

10 (24.4)

FE = 7.071;

Yes (n = 24)

11 (45.8)

0 (0.0)

13 (54.2)

P = 0.023

Lack of access to electronic database

Duration

Group I(n = 24

12 (50.0)

0 (0.0)

12 (50.0)

FE = 9.612;

Group II (n = 41)

25 (61.0)

8 (19.5)

P = 0.007

Setting

Academic (n= 33)

19 (57.6)

8 (24.2)

6 (18.2)

FE = 11.725;

Non-Academic (n = 32)

18 (56.3)

0 (0.0)

14 (43.7)

P = 0.002

EBM Trained

No (n = 41)

28 (68.3)

8 (19.5)

5 (12.2)

FE = 19.198;

Yes (n = 24)

9 (37.5)

0 (0.0)

15 (62.5)

P < 0.001

*FE = Fisher’s Exact Test; EBM – Evidence-based Medicine

Table 6b: Relationship between professional characteristics of the respondents and the perceived barriers to the use of EBM in

Paediatric Practice

Lack of skill to locate best research

Agreed

Undecided

Disagreed

Statistics

evidence

Duration

Group I(n = 24

14 (58.3)

0 (0.0)

10 (41.7)

FE = 14.312;

Group II (n = 41)

34 (82.9)

5 (12.2)

2 (4.9)

P < 0.001

Setting

Academic (n= 33)

28 (84.8)

3 (9.1)

2 (6.1)

χ2 = 6.853;

Non-Academic (n = 32)

20 (62.5)

2 (6.3)

10 (31.2)

P = 0.033

EBM Trained

No (n = 41)

34 (82.9)

5 (12.2)

2 (4.9)

FE = 14.312;

Yes (n = 24)

14 (58.3)

0 (0.0)

10 (41.7)

P < 0.001

Lack of skills for understanding statistics

Duration

Group I(n = 24

16 (66.7)

0 (0.0)

8 (43.3)

χ2 = 4.134*;

Group II (n = 41)

37 (90.2)

0 (0.0)

4 (9.8)

P = 0.042

Setting

Academic (n= 33)

21 (63.6)

2 (6.1)

10 (30.3)

χ2 = 5.277*;

Non-Academic (n = 32)

22 (68.7)

8 (25.0)

2 (6.3)

P = 0.022

EBM Trained

No (n = 41)

37 (90.2)

0 (0.0)

4 (9.8)

χ2 = 4.134*;

Yes (n = 24)

16 (66.7)

0 (0.0)

8 (33.3)

P = 0.042

Lack of incentives for using EBM in practice

Duration

Group I(n = 24

20 (83.3)

0 (0.0)

4 (16.7)

FE = 7.567;

Group II (n = 41)

22 (53.7)

8 (19.5()

11 (26.8)

P = 0.02

Setting

Academic (n= 33)

22 (66.7)

4 (12.1)

7 (21.2)

χ2 = 0.147;

Non-Academic (n = 32)

20 (62.5)

4 (12.5)

8 (25.0)

P = 0.929

EBM Trained

No (n = 41)

24 (58.5)

8 (19.5)

9 (22.0)

FE = 5.715;

Yes (n = 24)

18 (75.0)

0 (0.0)

6 (25.0)

P = 0.06

Fear of medicolegal tussles following EBM use

Duration

Group I(n = 24

8 (33.3)

0 (0.0)

16 (66.7)

FE = 8.791;

Group II (n = 41)

10 (24.4)

11 (26.8)

20 (48.8)

P = 0.013

Setting

Academic (n= 33)

10 (30.3)

8 (24.2)

15 (45.5)

χ2 = 3.395;

Non-Academic (n = 32)

8 (25.0)

3 (9.3)

21 (65.7)

P = 0.167

EBM Trained

No (n = 41)

13 (31.7)

11 (26.8)

17 (41.5)

FE = 11.727;

Yes (n = 24)

5 (20.8)

0 (0.0)

19 (79.2)

P = 0.003

*Chi- Square with Yate’s Correction; FE = Fisher’s Exact Test; EBM – Evidence-based Medicine

168

Table 6b shows that higher proportions of Group I

in a population of Iranian doctors. Although the rela-

respondents, respondents in non-academic settings and

tively high knowledge score observed in the present

respondents trained in EBM disagreed that lack of skills

study is reassuring, it may not be a perfect proof of deep

required to locate best research evidence is a barrier to

knowledge of EBM, given the low rate of training re-

EBM use (p< 0.001, p = 0.033 and p < 0.001 respec-

ported in the same population. Almost all the partici-

tively). Lack of skills for understanding statistics was

pants (92.3%) were willing to learn the skills required to

not perceived as a barrier to EBM by significantly

incorporate EBM into their routine clinical practice and

higher proportions of Group I respondents, respondents

this may be a reflection of the acceptability of EBM in

in the academic setting and respondents who were

clinical practice.

trained on EBM (p = 0.042, p = 0.022 and p = 0.042

respectively). In addition, a significantly higher propor-

The major perceived barriers to the use of EBM in the

tion of Group I respondents agreed that lack of incen-

present study included the lack of skills for understand-

tives for using EBM in practice was a barrier to the use

ing statistics, lack of skills for locating best research

of EBM (p = 0.02). Significantly higher proportions of

evidence and lack of incentives for using EBM. Interest-

respondents in Group I and those who were trained in

ingly, the fear of medicolegal tussles arising from the

EBM disagreed that there may be medicolegal tussles

practice of EBM was the least identified barrier. The

following EBM use (p = 0.013 and p = 0.003 respec-

lack of time was not perceived a leading barrier in the

present study unlike other previous studies in Jordan,

tively).

Sudan,

Saudi Arabia,

Norway,

and Sri Lanka.

This may reflect the heterogeneous nature of the partici-

pants in the various cited studies. Lower cadre doctors

Discussion

are likely to be busier in terms of clinical duties than the

higher cadre doctors, who are mostly involved in admin-

The present study revealed that almost all the respon-

istrative duties alongside academics with some involve-

dents (90.8%) used EBM in their clinical practices sug-

ment in clinical duties. Therefore, studies focused on

gesting a high level of acceptability of EBM in the

higher cadre doctors, like the present study, are not

population. This contrasts with 40% EBM use rate re-

likely to report lack of time or overwhelming workload

ported among doctors in Jordan and 5% each reported

as perceived barriers to the practice of EBM. However,

in Sudan and Jordan.

7,10

The observed difference may be

the challenge of lack of time, though coming a distant

explained in terms of differences in the degree of expo-

sixth position on a list of eight in the present study, may

sure to EBM during professional workshops and confer-

be addressed by incorporating research day-off on a

ences over time. However, the high use rate of EBM in

weekly or bi-weekly basis into the duty schedule for

the present study contrasted sharply with the low pro-

doctors at all levels.

portion of participants (36.9%) who had received train-

ing on EBM. This pattern was similar to the findings

Lack of understanding of statistics and lack of skills for

among a cohort of Nigerian specialist trainees, where

searching and locating best research evidence are train-

96.6% were familiar with EBM but only 38.8% had

ing issues which frequent Continuing Medical Pro-

been formally trained on EBM. This low rate of train-

grammes can address to a large extent. Access to inter-

ing was similar to 15% and 24% previously reported in

net EBM databases such as the Cochrane Library and

Sudan and Sri Lanka respectively.

10,12

This observation

the World Health Organisation Reproductive Health

may be related to the poor awareness of training oppor-

Library, may need to be improved, just as an earlier

study in Bosnia Herzegovina

tunities as currently provided by organisations such as

reported that 34.6% of

the various national and regional Cochrane Centres. It is

doctors did not know how to use the Cochrane database

plausible that paucity of funds and logistic supports may

similar to 38.5% observed in the present study. Institu-

actually limit the number of people that such orgnisa-

tions should be encouraged to subscribe to these data-

tions could admit for training at a time. Interestingly, the

bases and facilitate easy and hitch-free access and use by

higher cadre participants in the present study were four

doctors, particularly when they are not kept busy with

times more likely to have received training on EBM

clinical duties. Lack of evidence as a perceived barrier

compared to the junior cadre participants. The reason for

to the use of EBM had been reported by Al-Almaie in

Saudi Arabia. That may be related to the fact that the

this observation is obscure but it is plausible that the

senior participants had better access to training opportu-

Cochrane Database of Systematic Reviews display many

nities by virtue of hierarchy in the profession. The uni-

systemic reviews with inconclusive findings arising

versal recommendation of the participants in the present

from either small number of study participants or poor

study that training on EBM should be incorporated into

design of the primary randomized controlled trials.

postgraduate training curricula may provide a solution to

Therefore, this dearth of studies with conclusive find-

the lopsidedness in the pattern of EBM training in rela-

ings should not deter doctors from the use of the avail-

tion to the duration of practice.

able facts. In addition, this should also stimulate more

research in the form of well-designed, high quality ran-

Fourth-fifth of the participants in the present study had

domized controlled trials to answer some of the yet un-

high knowledge of what EBM entails and what purpose

answered research questions.

it serves. This is not surprising as the overall mean

knowledge score was 77.6% compared to 24% recorded

169

Al-Omari reported threat to clinical freedom as a per-

usually routinely addressed during EBM training and

ceived barrier to the use of EBM among doctors in Jor-

with practice, every researcher learns that the perceived

dan but that fear of medicolegal tussle surrounding the

barriers can be adequately overcome.

use of EBM did not appear strong in the present study.

This difference may be related to the relative rate of

medical litigations in each environment. Where treat-

ment protocols have been drawn for routine clinical

Conclusion

practice from the best quality evidences available, it

becomes difficult to defend any other clinical decision

In conclusion, the knowledge of this population of Nige-

aside the recommendations of the treatment protocol and

rian paediatricians about EBM was good but most of the

that may predispose to litigations. In Nigeria, the proc-

participants crave better understanding of the statistical

ess of generating treatment protocols for paediatric prac-

aspects of systematic reviews and EBM. The partici-

tice are still on-going, hence there is no pressure for

pants also recommended the inclusion of EBM training

mandatory adherence to specific practices yet. There-

in both the undergraduate and postgraduate medical cur-

fore, medicolegal issues may not yet be perceived as a

ricula as well as in Continuing Medical Education pro-

threat to the use of EBM.

grammes. With adequate training, the perceived barriers

to use of EBM will be addressed and appropriate solu-

The Group I participants (relatively higher in the hierar-

tions will be provided.

chy) and those who had received trainings on EBM were

The electronic method of gathering the responses of the

more likely to disagree with the perceptions of lack of

participants is acknowledged as a limitation to the study

computing skills, lack of access to electronic databases,

as it may explain the low response rate.

lack of skill to locate the best research evidence, lack of

skill for understanding statistics, lack of incentives and

fear of medicolegal tussles as barriers to EBM. This

Conflict of interest: None

observation may be explained in terms of the role of

Funding: None

experience on perceptions. The perceived barriers are

References

1. Oladapo OT. Research synthe-

7. Higgins J, Green S. Cochrane

12. Abeysena C, Jayawardana P,

sis: Implications for clinical

Handbook for Systematic Re-

Wickremasinghe R, Wickra-

practice in Nigeria. Interna-

views of Interventions. Coch-

mansinghe U. Evidence-based

tional J Multi disciplinary Re-

rane Book Series. Willey-

medicine knowledge, attitudes

search 2012; V(1&2): 4-6.

Blackwell. Chichester, West

and practices among doctors

2. Rosenberg WM, Sackett DL.

Sussex, England. 2008.

in Sri Lanka. J Evid Based

On the need for evidence based

8. Barghouti F, Halaseh L, Said T,

Med 2010; 3(2): 83-87.

medicine. Therapie 1996; 51

Mousa AH, Dabdoub A. Evi-

13. Rashidbeygi M, Sayehmiri K.

(3): 212-217.

dence-based medicine among

knowledge and attitude of

3. Collins JA. Clinical research

Jordanian family physicians:

physicianstowarss evidence

evidence and clinical practice.

awareness, attitude and knowl-

based medicine in Ilam, Iran.

Hum Reprod 1997; 12(9): 1847

edge. Can Fam Physician

Iran Red Crescent J 2013; 15

-1850.

2009; 55(7): e6-e13 .

(9): 798-803.

Mahmoud AO, Ayanniyi AA,

9. Zeidan AZ, Behairy MM.

14. Al-Musa HM. Knowledge,

Lawal A, Omolase CO,

Knowledge, attitude and prac-

perceptions, attitude and edu-

Ologunsua Y, Samaila E. Sur-

tice of Evidence-based medi-

cational needs of physicians to

vey of medical specialists on

cine among residence doctors

evidence based medicine in

their attitudes to and resources

in Sudan. Sudan J Med Sci

South-Western Saudi Arabia.

for health research in Nigeria.

2010; 5(3): 2007-2012.

Saudi Med J 2010; 31(3): 308

Ann Afr Med 2011; 10(2): 144-

10. Al Omari M, Khader Y, Jadal-

-312.

149 .

lah K, Dauod AS, Al-Shdifat

15. Ulvenes LV, Aasland O,

Mahmoud AO, Ayanniyi AA,

AA, Khasawneh NM. Evidence

Nylenna M, Kristianseh IS.

Lawal A, Omolase CO,

-based medicine among hospi-

Norwegian physicians’ knowl-

Ologunsua Y, Samaila E. Per-

tal doctors in Jordan: aware-

edge of and opinions about

ceptions of Nigerian medical

ness, attitude and practice. J

evidence-based medicine:

specialists on research. J Public

Eval Clin Pract 2009; 15(6):

cross-sectional study. PLoS

Health Afr 2011; 2(1): e1. Doi:

1137-1141.

One 2009; 4(11): e7828. Doi:

10.4081/jphia.2011.e1

11. Fawole A, Oladapo O, Enahoro

10.1371/

Sackett DL, Rosenberg WM.

F, Akande E. Acceptance of

journal.pone.0007828.

Th need for evidence-based

evidence-based reproductive

medicine. J R Soc Med 1995;

health care among postgraduate

88(11): 620-624.

specialist trainees in Nigeria. Int J

GynaecolObstet 2008; 102(1): 3- 7.

170

16. Mahmić - KaknjoM, Kadić D,

17. Al-Almaie SM, Al-Baghli N.

Hodžic H, Spahić - Sarajlić S,

Barriers facing physicians phy-

Hadžic E, Ademović E. Aware-

sicians practicing evidence

ness, knowledge, use and atti-

based medicine in Saudi Ara-

tude towards evidence based

bia. J ContinEduc Health Prof

medicine in a developing coun-

2004; 24(3): 163-170.

try: Survey of physicians in a

canton in Bosnia and Herzego-

vina . Croat Med J 2015; 56(6):

558-566.