535507

research-article2014
JHLXXX10.1177/0890334414535507Journal of Human LactationPound et al

Original Research
Journal of Human Lactation

Breastfeeding Knowledge, Confidence,

1­–12
© The Author(s) 2014
Reprints and permissions:
Beliefs, and Attitudes of Canadian sagepub.com/journalsPermissions.nav
DOI: 10.1177/0890334414535507

Physicians jhl.sagepub.com

Catherine M. Pound, MD1, Kathryn Williams, BSc, MS2, Renee Grenon, MA, PhD(c)2,
Mary Aglipay, MSc2, and Amy C. Plint, MD, MSc1,2

Abstract
Background: Physicians’ attitudes and recommendations directly affect breastfeeding duration. Yet, studies in many nations
have shown that physicians lack the skills to offer proper guidance to breastfeeding mothers.
Objective: This study aims to assess breastfeeding knowledge, confidence, beliefs, and attitudes of Canadian physicians.
Methods: A breastfeeding questionnaire was developed and piloted prior to study enrollment. These questionnaires were
sent to 1429 pediatricians (PED), 1329 family physicians (FP), and final-year pediatric and final-year family medicine residents
(PR and FMR).
Results: The analysis included 397 PED, 322 FP, 17 PR, and 44 FMR who completed the questionnaire. Mean overall correct
knowledge score was 67.8% for PED, 64.3% for FP, 72.7% for PR, and 66.8% for FMR. Two hundred eighty-five PED (74.2%),
228 FP (73.1%), 7 PR (41.2%), and 21 FMR (53.8%) felt confident with their breastfeeding counseling skills. Less than half
(49.6% of PED and 45.4% of FP) believed that evaluating breastfeeding was a primary care physician’s responsibility, and few
PED or FP (5.1% and 11.3%) routinely observed breastfeeding in mother-infant pairs.
Conclusion: Several areas of potential deficits were identified in Canadian physicians’ breastfeeding knowledge. Physicians
would benefit from greater education and support, to optimize care of infants and their mothers.

Keywords
attitudes, beliefs, breastfeeding, confidence, education, humans, infant, knowledge, medicine, newborn, physician’s role, practice

Well Established in many countries have shown that physicians and residents
lack skills to offer proper guidance to lactating mothers.4-10
Physicians’ attitudes and recommendations are known to directly In Canada, the learning of specific breastfeeding skills
affect breastfeeding duration. Studies in many nations have shown and the management of common breastfeeding problems
that physicians often lack skills and/or knowledge to assist and do not appear among the programs’ regulatory bodies’ for-
support breastfeeding mothers. mal learning objectives,11,12 suggesting that breastfeeding
counseling skills are not formally taught during residency
Newly Expressed training.
We undertook a national assessment of Canadian physi-
Several areas of potential deficits were identified in Canadian cians caring for infants and their mothers (pediatricians
physicians’ breastfeeding knowledge, including appropriate [PED], family physicians [FP], final-year pediatric residents
breastfeeding techniques, latch, and recommendations pertain-
ing to milk supply. Physicians would benefit from greater educa- 1
Children’s Hospital of Eastern Ontario, Ottawa, ON, Canada
tion and support, to optimize care of infants and their mothers. 2
Children’s Hospital of Eastern Ontario Clinical Research Unit, Ottawa,
ON, Canada

Background Date submitted: January 14, 2014; Date accepted: April 22, 2014.

The Canadian Pediatric Society (CPS), the American Academy Corresponding Author:
Catherine M. Pound, MD, Department of Pediatrics, Division of Pediatric
of Pediatrics (AAP), and the American Academy of Family Medicine, Children’s Hospital of Eastern Ontario, 401 Smyth Road,
Physicians (AAFP) promote the assessment and management Ottawa, ON, K1H 8L1, Canada.
of breastfeeding issues as physician responsibilities.1-3 Studies Email: [email protected]

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

2 Journal of Human Lactation 

[PR], and final-year family medicine residents [FMR]) to Lactation Consultants further reviewed the questionnaire. As
examine their breastfeeding knowledge, confidence, beliefs, no standardized questionnaire was found in the literature,
and attitudes. To our knowledge, this is the most extensive questions were developed based on expected knowledge of
study of the topic to be performed in Canada. the primary care physician detailed in the AAP Position
Statement on “Breastfeeding and the Use of Human Milk,”1
adapted from the Registered Nurses’ Association of Ontario
Methods Self-Learning Module on Breastfeeding14 and from prior
Population and Survey Distribution studies.6,10 Questions from more recent studies15 could not be
included as publication date was too late for incorporation
We conducted a paper-based, mailed survey of 1429 PED into our study. Permissions were obtained from all appropri-
and 1329 FP across Canada (all provinces and territories) ate sources.6,10,14 Twenty-three physicians and residents
between October 2010 and January 2011. Names and contact piloted the survey, which was revised based on their feed-
information of all licensed PED and FP were obtained from back. The final survey consisted of 50 multiple-choice ques-
all provincial registries except for Quebec. The Collège des tions covering demographics and 4 main domains:
Médecins du Québec (CMQ) did not release their physician knowledge, confidence, beliefs, and attitudes (17, 4, 5, and 8
list but agreed to share names and contact information of 400 questions, respectively). All questions were given equal
PED and 400 FP they randomly selected. A computerized weight for scoring purposes. The knowledge part of the ques-
program randomly selected participants in proportion to the tionnaire included 4 multiple-choice questions, 8 “true/
number of physicians working in each province, except for false/I don’t know” questions, and 5 “yes/no/I don’t know”
Quebec, where all randomly selected 800 physicians were questions. The confidence part of the questionnaire used 1
mailed a questionnaire. Also, due to the limited number of “yes/no” question and 3 multiple-choice answers, the belief
physicians working in the northern territories (Nunavut, part used 3 “yes/no” questions and 2 multiple-choice
Yukon, and Northwest Territories), all PED and FP in these answers, and the attitude part used 3 “yes/no” questions and
areas were surveyed. A modified Dillman method13 was 5 multiple-choice answers (not comfortable, somewhat
used; nonresponders received up to 3 surveys sent at regular uncomfortable, somewhat comfortable, very comfortable;
intervals via regular mail. Only physicians who provided pri- never or almost never, on some visits, on most visits, almost
mary care for infants were eligible, and as such, pediatric or almost always). The confidence section of the question-
subspecialists were excluded. Physicians whose type of prac- naire investigated self-rated confidence in one’s breastfeed-
tice clearly excluded infants and primary care (geriatrics, ing counseling skills. Overall confidence was defined as
anesthesia, palliative care, etc) were excluded. Surveys were “confidence in teaching mothers how to breastfeed and
available in both French and English. address breastfeeding-related problems.” In the belief sec-
We also conducted an electronic survey of final-year PR tion, physicians’ thoughts in relation to their role in breast-
and final-year FMR from all programs over the same time feeding support and promotion were explored, whereas
frame. As most trainees’ names and contact information can- physicians’ usual clinical practice and outlook on various
not be publicly accessed, we asked program directors to dis- breastfeeding scenarios were explored in the attitude
tribute a letter to their eligible residents. Interested residents section.
communicated directly with our research assistant. Because We a priori defined an overall knowledge score of 70% as
no national public list of residents exists, we estimated the acceptable, as this is the minimum score that pediatricians
number of residents enrolled in their final year of residency must achieve in their specialty examination to receive Royal
during the study period, based on data from the Canadian College of Physicians and Surgeons of Canada certification
Residency Matching System. Ninety to 100 PR and 800 to (M. Jabbour, MD, vice-chief/chair, Department of Pediatrics,
875 FMR were likely eligible. associate professor, Pediatrics and Emergency Medicine,
The study was approved by the Children’s Hospital of University of Ottawa, personal communication, May 14,
Eastern Ontario Research Ethics Board, and all procedures 2013). There is no minimum score for family physicians on
followed were in accord with the ethical standards of the certification examination; within each examination adminis-
board. tration, a criterion candidate group is selected whose perfor-
mance is the standard against which all candidates are
Survey Instrument compared.16 Since both groups of physicians care for the
same infant population, we felt that the minimum acceptable
Since a review of the breastfeeding literature failed to yield score should be the same.
an existing questionnaire of large enough breadth to satisfac-
torily examine our 4 domains (knowledge, confidence,
Sample Size
beliefs, and attitudes), the study principal investigator, in
conjunction with the study team, developed the survey We based our sample size calculation on the results of our
instrument. Two practicing International Board Certified pilot study, in which the mean percentage of correct answers

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

Pound et al 3

in the knowledge section was 65%, with a standard deviation of differing scales so as to keep scales homogeneous. In total,
of 12%. A sample size of 385 allowed for an estimate of the 1 question was removed from the confidence domain, none
mean percentage score in the knowledge section with a 95% for knowledge, 3 for attitudes, and 2 for beliefs. Excellent
confidence interval (CI) of width ± 1.2%. Assuming a 30% reliability was found for confidence (n = 761, Cronbach’s
response rate, we had planned to approach 1284 PED and alpha = 0.81), whereas acceptable reliability was demon-
1284 FP. However, since the CMQ would release contact strated for knowledge (n = 625, Cronbach’s alpha = 0.62)
information for only a predetermined number of physicians, and attitudes (n = 506, Cronbach’s alpha = 0.60). The belief
we overestimated the number needed from that province and domain has a relatively low reliability value (n = 755,
asked to have 800 names released, as some physicians would Cronbach’s alpha = 0.21).
not meet eligibility criteria. Also, given the very small num-
ber of physicians practicing in the Northern Territories, we
Knowledge
elected to approach them all. In the end, 1429 PED and 1329
FP were approached. The average overall knowledge score of practicing physi-
cians was lower than our predefined acceptable score of
70%. The overall correct knowledge score for PED was
Statistical Analysis 67.8%, whereas that for FP was 64.3%. Pediatric residents
Breastfeeding knowledge was calculated as a score out of 100 scored 72.7%, whereas FMR scored 66.8%. Table 2 summa-
based on the percentage of correct responses from the breast- rizes physicians’ and residents’ results on the questionnaires’
feeding knowledge domain. Descriptive statistics were used to knowledge section. Univariate testing demonstrated that
summarize the survey respondents’ demographic characteris- knowledge score was associated with several demographic
tics, breastfeeding confidence and beliefs, and breastfeeding factors (Table 1).
attitudes. Cronbach’s alpha values were calculated for each
domain of the questionnaire. Univariate associations between Confidence, Beliefs, and Attitudes
knowledge and demographics were assessed using t tests and
1-way analyses of variance (ANOVAs). Multiple linear regres- Two hundred eighty-five PED (74.2%) and 228 FP (73.1%)
sion was used to identify demographic factors that were inde- reported feeling overall confident in teaching mothers to
pendently associated with the overall knowledge score. All breastfeed and addressing breastfeeding-related problems.
demographic variables surveyed were included in the multi- Few PED and FP (5.1% and 11.3%, respectively) routinely
variate model. Missing data were handled via listwise dele- observed breastfeeding at least once in every breastfeeding
tion. The identification of demographic factors associated with mother-infant pair, and a large number of PED and FP (56.6%
knowledge was the primary goal of the modeling, so only and 29.9%, respectively) reported keeping formula samples
main effects (no interactions) were assessed. in their office. Table 3 summarizes self-reported confidence,
A P value < .05 was considered statistically significant. beliefs, and attitudes of physicians and residents regarding
The analyses were performed using SPSS (version 19.0; various aspects of breastfeeding.
IBM Corp, Armonk, New York, USA) and SAS 9.3 (SAS
Institute Inc, Cary, North Carolina, USA). Comparison between FP and PED
The overall knowledge score for PED was higher than that
Results for FP (P = .007). However, more PED than FP believed that
formula was equivalent to breast milk (P < .001). More FP
Respondents reported directly observing breastfeeding at least once in
We identified 3395 PED and 38 223 FP from the provincial every breastfeeding mother (P = .004), and more PED kept
and territorial physician lists. Response rate was 38.3% for formula samples in the office (P < .001). Other confidence,
PED and 24.2% for FM. Overall, 719 physicians represent- beliefs, and attitudes of PED and FP were not significantly
ing all provinces and territories were retained for the analysis different.
(397 PED and 322 FM) (Figures 1-3; Table 1). Sixty-four Multiple linear regression identified several demographic
residents contacted the research assistant, and 61 returned factors that were significantly associated with knowledge for
surveys (17 PR and 44 FMR). The majority of participants practicing physicians (Table 4). On average, female physi-
were female and between the ages of 30 and 50. cians scored 5.0 points higher (on a 100-point scale) com-
pared to male physicians (95% CI, 2.4-7.5; P < .001).
Compared to older physicians, physicians between the ages
Questionnaire of 30 and 50 scored higher by 3.6 points (95% CI, 0.2-7.1;
Cronbach’s alpha values were calculated for each domain of P = .04). Compared to Ontario, the Maritime Provinces and
the questionnaire. Yes/no questions were removed for British Columbia had higher knowledge scores, scoring on
Cronbach’s alpha calculation from domains with questions average 5.2 (95% CI, 0.6-9.8; P = .03) and 4.3 (95% CI, 0.3-8.3;

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

4 Journal of Human Lactation 

Figure 1.  Pediatrician Mail-out.

3395 Canadian pediatricians idenfied

741 pediatricians removed
(no available address, out of
country / out of province,
iden
fied as subspecialists)
2654 pediatricians remained

1463 pediatricians randomly selected

31 pediatricians removed
(28 missing addresses not
previously identified, 2 duplicate
addresses, 1 out of province
address not previously identified)
1429 quesonnaires sent

8 envelopes returned unopened

(moved / wrong address)

1421 quesonnaires delivered

588 quesonnaires returned 40 blank/incomplete

(3 saw no infants, 5 no longer in
prac
ce / re
red, 2
subspecialists, 27 gave no reason,
3 incomplete)
548 completed quesonnaires

151 quesonnaires excluded

(pediatric subspecialists)
397 quesonnaires in final analysis

P = .04) points higher, respectively. Physicians whose prac- level was relatively high, self-reported confidence was much
tice had 50% or more patients younger than 1 year scored on lower when participants were asked about specific breast-
average 7.9 points higher on the knowledge tool (95% CI, feeding counseling skills, suggesting that physicians may
3.6-12.1; P < .001). Compared to physicians with no children, have difficulty self-assessing as they may not be aware of
physicians who had breastfed children scored 10 points higher what breastfeeding counseling entails. Physicians’ attitudes
(95% CI, 6.4-13.6; P < .001). Compared to physicians with toward a breastfeeding infant were found to be quite posi-
no children, having non-breast-fed children was not associ- tive, however, a much smaller number felt comfortable
ated with a higher knowledge score (average increase in watching a toddler breastfeed, despite the World Health
score = 4.5; 95% CI, –2.7 to 11.7; P = .22). Breastfeeding cer- Organization supporting breastfeeding well into toddler-
tification was associated with an 11.5-point increase (95% CI, hood.17 Finally, female sex, age between 30 and 50 years,
1.7-21.2; P = .02). The demographic variables in our model percentage of practice younger than 1 year, and personal
accounted for 14.1% of the variation in knowledge score. breastfeeding experience were all positively associated with
knowledge.
This survey identified multiple knowledge deficits. These
Discussion included recommending inappropriate breastfeeding tech-
We found that Canadian physicians’ breastfeeding knowl- niques (such as timed feedings), incorrectly believing that
edge was suboptimal. Although physicians’ overall comfort increasing maternal milk intake increases breast milk supply,

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

Pound et al 5

Figure 2.  Family Physician Mail-out.

38 223 Canadian family physicians

identiied 4376 family physicians
removed
(no available address, out of
country / out of province,
ield
of work that excludes pediatric
population; e.g. geriatrician)
33 847 family physicians remained

1362 family physicians randomly

33 family physicians removed
selected
(10 missing addresses not
previously identi
ied, 2 out of
province address not previously
identi
ied, 21 with specialties
clearly excluding pediatric
patients)
1329 questionnaires sent
15 envelopes returned
unopened
(moved / wrong address)
1314 questionnaires delivered

356 questionnaires returned 34 questionnaires excluded

(1
illed by nurse, 5 saw no
infants, 3 no longer in practice /
retired, 25 incomplete)

322 completed questionnaires

Figure 3.  Resident Electronic Mail-out.

17 pediatric postgraduate training 17 family medicine postgraduate

programs identiied training programs identiied

17 program directors approached 17 program directors approached

for letter distribution to their inal for letter distribution to their inal
year residents year residents

17 pediatric residents contacted the 44 family medicine residents

study coordinator and completed contacted the study coordinator and
the survey completed the survey

failing to identify characteristics of a successful latch, not unaware that early formula introduction is a major risk factor
recognizing breastfeeding contraindications, and being for breastfeeding failure.18,19 It is interesting that about 75%

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

6 Journal of Human Lactation 

Table 1.  Demographics of Responding Physicians and Residents.

Pediatric Family
Family Physicians, Residents, Medicine
Pediatricians, n (%) n (%) n (%) Residents, n (%)

Total No. 397 322 17 44

Age, ya
  < 30 2 (0.5) 16 (5.0) 5 (29.4) 25 (56.8)
 30-50 203 (51.1) 150 (46.6) 12 (70.6) 19 (43.2)
  > 50 191 (48.2) 156 (48.4) 0 (0) 0 (0)
Sexa
 Male 169 (42.6) 145 (45.3) 5 (29.4) 8 (18.2)
 Female 228 (57.4) 175 (54.7) 12 (70.6) 36 (81.8)
Lived in Canada
 Always 225 (56.8) 202 (63.3) 14 (82.4) 32 (72.7)
  > 20 years but not always 116 (29.3) 65 (20.4) 2 (11.8) 5 (11.4)
  11-20 years 20 (5.1) 26 (8.2) 1 (5.9) 1 (2.3)
  5-10 years 22 (5.6) 19 (6.0) 0 (0) 5 (11.4)
  < 5 years 13 (3.3) 7 (2.2) 0 (0) 1 (2.3)
Province/territory of worka
  Atlantic provinces 36 (9.2) 21 (6.6) 2 (11.8) 8 (18.2)
 Quebec 97 (24.7) 63 (19.7) 2 (11.8) 8 (18.2)
 Ontario 160 (40.8) 126 (39.4) 11 (64.7) 24 (54.5)
  Prairie provinces 62 (15.8) 44 (13.8) 2 (11.8) 4 (9.1)
  British Columbia 34 (8.7) 46 (14.4) 0 (0) 0 (0)
  Nunavut, Yukon, Northwest Territories 3 (0.8) 20 (6.3) NA NA
Residency training in Canada
 Yes 343 (88.2) 258 (81.4)  
Work location
 Rural 61 (16.2) 120 (38.2)  
 Urban 316 (83.8) 194 (61.8)  
Type of work
 Clinic 197 (50.6) 246 (77.4)  
  Community hospital 61 (15.7) 30 (9.4)  
  Teaching hospital 113 (29.0) 11 (3.5)  
 Other 18 (4.6) 31 (9.7)  
Percentage of children < 1 year of age in practicea
 0-10 43 (11.1) 236 (76.1)  
 11-25 171 (44.3) 59 (19.0)  
 26-50 106 (27.5) 15 (4.8)  
 51-75 46 (11.9) 0 (0)  
  > 75 20 (5.2) 0 (0)  
Years in practicea
 0-5 77 (19.9) 52 (16.5)  
 6-10 40 (10.4) 36 (11.4)  
 11-15 51 (13.2) 32 (10.1)  
  > 15 218 (56.5) 196 (62.0)  
Breastfeeding learning
  Own experiencea 255 (64.2) 204 (63.4) 7 (41.2) 20 (45.5)
  Medical school 117 (29.5) 143 (44.4) 7 (41.2) 25 (56.8)
 Residency 196 (49.4) 110 (34.2) 15 (88.2) 30 (68.2)
  Self-directed learninga 195 (49.1) 150 (46.6) 6 (35.3) 20 (45.5)
 Othera 74 (18.6) 77 (23.9) 1 (5.9) 8 (18.2)
  No breastfeeding knowledgea 1 (0.3) 3 (0.9) 0 (0) 0 (0)

(continued)

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

Pound et al 7

Table 1. (continued)

Pediatric Family
Family Physicians, Residents, Medicine
Pediatricians, n (%) n (%) n (%) Residents, n (%)
a
Having 1 or more children
 Yes 335 (87.7) 265 (83.6) 8 (50.0) 17 (38.6)
Self/partner with breastfeeding experiencea
 Yes 312 (94.0) 237 (96.0) 8 (100.0) 17 (100.0)
No. months own child was breastfeda
  ≥ 12 114 (50.2) 143 (53.6) 4 (57.1) 9 (52.9)
  < 12 93 (41.0) 114 (42.7) 3 (42.9) 8 (47.1)
  No breastfeeding 20 (8.8) 10 (3.7) 0 (0) 0 (0)
Certification in breastfeeding support
 Yes 6 (1.5) 4 (1.3) 1 (5.9) 1 (2.3)

Abbreviation: NA, not applicable.

a
P < .01 for association with knowledge score among pediatricians and family physicians.

Table 2.  Breastfeeding Knowledge Score.

Respondents with Correct Answer, n (%)

Family Pediatric Family Medicine

Scenario Correct Answer Pediatricians Physicians Residents Residents

Overall score 67.8% 64.3% 72.7% 66.8%

For exclusively breastfed baby, No 276/380 (72.6) 224/311 (72.0) 11/17 (64.7) 27/44 (61.4)
otherwise healthy, birth weight
not regained by 2 weeks, is
first recommendation formula
supplementation?
Do you routinely recommend No 299/384 (77.9) 232/307 (75.6) 12/17 (70.6) 37/44 (84.1)
supplementing with formula
if mother feels milk supply
inadequate?
Do you routinely give glucose No 347/387 (89.7) 253/307 (82.4) 15/17 (88.2) 34/44 (77.3)
water or formula before
mother’s milk comes in?
Is formula feeding in first Yes 184/382 (48.2) 203/315 (64.4) 10/17 (58.8) 26/44 (59.1)
few weeks risk factor for
breastfeeding failure?
Do you routinely recommend that No 159/385 (41.3) 141/308 (45.8) 2/17 (11.8) 14/44 (31.8)
babies breastfeed each side for
15-20 minutes every 3 hours?
First thing to do when mother Assess position and latch 328/376 (87.2) 257/302 (85.1) 17/17 (100.0) 42/44 (95.5)
complains of sore nipples is:
All listed answers are signs that No part of the areola can 117/367 (31.9) 102/300 (34.0) 8/17 (47.1) 16/43 (37.2)
baby is latched on properly be seen
except:
Mother complains that 6-week-old Baby requires more milk 256/374 (68.4) 173/305 (56.7) 12/17 (70.6) 27/44 (61.4)
infant has been breastfeeding because growing and
hourly for a day or 2. You tell frequent breastfeeding
her: increases milk supply

(continued)

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

8 Journal of Human Lactation 

Table 2. (continued)

Respondents with Correct Answer, n (%)

Family Pediatric Family Medicine

Scenario Correct Answer Pediatricians Physicians Residents Residents
Otherwise healthy 5-day-old More frequent breastfeeding 338/380 (91.4) 244/307 (79.5) 17/17 (100.0) 39/44 (88.6)
breastfeeding infant admitted sessions and teach mother
to hospital with jaundice. After how to use breast pump
initiating phototherapy, what do
you recommend?
Moderate exercise decreases False 322/390 (82.6) 263/318 (82.7) 14/17 (82.4) 33/44 (75.0)
quality and quantity of breast
milk
Increasing mother’s milk intake False 107/390 (27.4) 93/314 (29.6) 6/17 (35.3) 19/44 (43.2)
increases breast milk production
Breastfeeding decreases incidence True 372/391 (95.1) 310/319 (97.2) 17/17 (100.0) 40/44 (90.9)
of many infectious diseases
Breastfeeding decreases risk of True 285/386 (73.8) 230/317 (72.6) 16/17 (94.1) 32/44 (72.7)
SIDS
Breastfeeding is contraindicated in False 236/387 (61.0) 145/317 (45.7) 9/17 (52.9) 19/44 (43.2)
mothers with hepatitis C
Breastfeeding is contraindicated in True 311/388 (80.2) 144/317 (45.4) 17/17 (100.0) 36/44 (81.8)
mothers with HIV in Canada
Breastfeeding decreases the risk True 272/390 (69.7) 246/319 (77.1) 14/17 (82.4) 35/44 (79.5)
of ovarian and breast cancers in
mothers
Breastfeeding is safe to continue in True 215/388 (55.4) 156/317 (49.2) 13/17 (76.5) 23/44 (52.3)
mothers with herpes simplex on
1 breast if baby breastfeeds only
from other breast

Abbreviations: HIV, human immunodeficiency virus; SIDS, sudden infant death syndrome.

Table 3.  Confidence, Beliefs, and Attitudes.

Pediatricians, Family Pediatric Family Medicine

Question n (%) Physicians, n (%) Residents, n (%) Residents, n (%)
Confidence
  Overall confidence in breastfeeding 285/384 (74.2) 228/312 (73.1) 7/17 (41.2) 21/44 (53.8)
counseling skills
  Very comfortable assessing baby’s latch 128/389 (32.9) 103/313 (32.9) 3/17 (17.6) 6/44 (13.6)
  Very comfortable assessing milk transfer 136/391 (34.8) 103/314 (32.8) 4/17 (23.5) 9/44 (20.5)
  Very comfortable teaching mothers breast 87/391 (22.3) 82/317 (26.9) 6/17 (35.3) 9/44 (20.5)
pump use
Beliefs
  Believe evaluation of breastfeeding is 191/385 (49.6) 142/312 (45.5) 13/17 (76.5) 29/44 (65.9)
responsibility of child’s primary physician
in first 3 to 5 days after birth
  Believe formula is nutritionally equivalent 132/380 (34.7) 70/314 (22.3) 5/17 (29.4) 10/44 (22.7)
to breast milk
  Believe physician has influence on mother’s 355/389 (91.3) 291/318 (91.5) 17/17 (100.0) 43/44 (97.7)
decision to breastfeed

(continued)

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

Pound et al 9

Table 3. (continued)

Pediatricians, Family Pediatric Family Medicine

Question n (%) Physicians, n (%) Residents, n (%) Residents, n (%)
  Believe it is always or almost always 67/384 (17.4) 72/317 (22.7) 2/17 (11.8) 7/44 (15.9)
realistic for working mothers to continue
breastfeeding
  Believe their residency training had 288/388 (74.2) 222/314 (70.7) 10/17 (58.8) 30/44 (68.2)
prepared them poorly or somewhat
poorly to support breastfeeding mothers
Attitudes
  Very comfortable witnessing mothers 352/390 (90.3) 289/319 (90.6) 16/17 (94.1) 31/44 (70.5)
breastfeeding infant in the office
  Very comfortable witnessing mothers 214/390 (54.9) 188/318 (59.1) 8/17 (47.1) 19/44 (43.2)
breastfeeding toddler in the office
  Always or almost always discuss 65/386 (16.8) 150/309 (48.5) 4/17 (23.5) 26/44 (59.1)
breastfeeding prior to birth of child
  Always or almost always ask about 273/387 (70.5) 202/311 (65.0) 7/17 (41.2) 26/44 (59.1)
breastfeeding in infant’s first year of life
  Directly observe breastfeeding at least 20/388 (5.1) 35/311 (11.3) 1/17 (5.9) 2/44 (4.5)
once in every breastfeeding mother-infant
pair
  Keep formula samples in the office (or 213/386 (56.6) 93/312 (29.9) 5/17 (29.4) 10/44 (22.7)
preceptor’s)
  Have formula advertisement in the office 77/378 (20.4) 50/313 (16.0) 8/17 (47.1) 14/44 (31.8)
(or preceptor’s)
  Keep breastfeeding pamphlets and 196/374 (52.4) 168/311 (54.0) 5/17 (29.4) 13/44 (29.4)
brochures in the office (or preceptor’s)
and give to breastfeeding mothers

Table 4.  Adjusted Model of the Overall Knowledge Score.a

Regression Coefficientb
Factor n (95% CI) P Value
Occupational group
  General pediatrician 300 2.27 (–0.50-5.03) .11
  Family physician 370 Reference  
Sex
 Female 386 4.95 (2.35-7.54) < .001
 Male 284 Reference  
Age
  ≤ 30 years 17 6.27 (–2.69-15.22) .17
  30-50 years 333 3.64 (0.17-7.12) .04
  > 50 years 320 Reference  
Lived in Canada
 Always 400 1.68 (–1.19-4.55) .25
  Not always 270 Reference  
Province
  Newfoundland, Prince Edward Island, Nova 54 5.18 (0.59-9.76) .03
Scotia, New Brunswick
 Quebec 147 –0.66 (–4.04-2.72) .70
  Manitoba, Saskatchewan, Alberta 99 –0.60 (–4.24-3.03) .75
  British Columbia 73 4.30 (0.27-8.33) .04

(continued)

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

10 Journal of Human Lactation 

Table 4. (continued)

Regression Coefficientb
Factor n (95% CI) P Value
  Nunavut, Yukon, Northwest Territories 23 4.90 (–2.09-11.89) .17
 Ontario 274 Reference  
Percentage of practice younger than 1 year
  > 50% 605 7.87 (3.64-12.10) < .001
  ≤ 50% 65 Reference  
Residency training in Canada
 Yes 567 0.26 (–3.52-4.03) .89
 No 103 Reference  
Rural or urban practice
 Rural 173 0.78 (–2.24-3.80) .61
 Urban 497 Reference  
Type of practice
 Community 88 1.27 (–2.51-5.05) .51
 Teaching 116 0.78 (–2.83-4.39) .67
 Other 44 –1.37 (–6.30-3.56) .59
 Clinic 422 Reference  
Years in practice
 0-5 119 1.22 (–3.16-5.60) .58
 6-10 74 2.19 (–2.37-6.75) .35
 11-15 79 1.62 (–2.82-6.06) .47
  > 15 398 Reference  
Personal breastfeeding experience
  No, has children 24 4.49 (–2.68-11.67) .22
  Yes, has children 506 9.99 (6.38-13.60) < .001
 Unknown 44 1.73 (–4.06-7.52) .56
  Does not have children 96 Reference  
Breastfeeding certification
 Yes 10 11.48 (1.74-21.23) .02
 No 660 Reference  

Abbreviation: CI, confidence interval.

a
The overall knowledge score was expressed as a percentage (n = 670).
b
Each regression coefficient indicates the additional % knowledge added by the corresponding factor to the reference mean score of 48.3% (constant
mean score of 48.3 not explained by the factors in the model) after adjusting for the other factors. Adjusted R2 = 0.14.

of physicians reported feeling overall comfortable address- an alarmingly low number of participants reported observing
ing breastfeeding problems, whereas only 41% of PR and breastfeeding at least once in their young patients, as this
54% of FMR reported the same. It is important to remember should be done for every infant in the first few days of life.1
that, although attending physicians were randomly selected, These results are concordant with those of previous studies,
the resident group self-selected. One would expect those showing that breastfeeding mothers receive minimal help
self-selected residents to be more interested in breastfeeding and support from physicians.20,21 A significant proportion of
than their peers and therefore more likely to be comfortable PED and FP reported keeping formula samples and formula
with the topic. The difference in scores between the 2 groups advertisements in their office, a practice shown to negatively
could suggest that breastfeeding skills are primarily learned affect breastfeeding duration.22
with clinical and life experience, rather than in residency. In keeping with other studies conducted world-
Indeed, most physicians and residents felt that residency wide,6,9,10,23,24 female sex and personal breastfeeding experi-
training had prepared them poorly or somewhat poorly to ence were associated with higher overall breastfeeding
appropriately support breastfeeding mothers. A considerable knowledge scores. Age 30 to 50 years was also associated
number of participants believed that breast milk and formu- with higher overall knowledge score, possibly reflecting a
las are nutritionally equivalent, despite unambiguous evi- group of physicians more likely to have recently had their
dence that formula is inferior.1 From an attitude perspective, own breastfeeding children.

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

Pound et al 11

The overall score of PED was slightly higher than that of Declaration of Conflicting Interests
FP but remained below our acceptable level of 70%. The authors declared no potential conflicts of interest with
Pediatricians reported keeping formula samples in the office respect to the research, authorship, and/or publication of this
more often than FP did. This finding, not previously reported, article.
may be because PED sometimes care for more complex or
medically fragile infants, with more significant feeding dif- Funding
ficulties, and in whom formula supplementation may at times The authors disclosed receipt of the following financial support for
be necessary. This may also be due to formula companies the research, authorship, and/or publication of this article: This
potentially targeting pediatricians more often. It is interest- study was funded by the Children’s Hospital of Eastern Ontario
ing that FP reported observing breastfeeding more often than Research Institute.
PED did. The reason for this is unclear but may relate to the
nature of the family physician’s relationship with the mother. References
Family physicians being typically involved in mothers’ care 1. American Academy of Pediatrics, Section on Breastfeeding.
prenatally, they may have a stronger bond with her and feel Breastfeeding and the use of human milk. Pediatrics.
more comfortable observing her breastfeed. 2012;129(3):e827-e841.
Study limitations include a low overall response rate. High 2. Pound CM, Unger S. Canadian Pediatric Society, Hospital
response rates are typically difficult to obtain from physi- Pediatrics Section, Nutrition and Gastroenterology Committee.
cians.25 Nevertheless, our target size was reached for PED The Baby-Friendly Initiative: Promoting, supporting and protect-
and almost reached for FP. Physicians interested in breast- ing breastfeeding. Paediatr Child Health. 2012;17(6):317-321.
feeding were probably more likely to complete the question- 3. American Academy of Family Physicians, Advisory Committee
naire, and therefore, our sampled population may not be on Breastfeeding. Breastfeeding, family physician supporting
(position paper). http://www.aafp.org/about/policies/all/
representative of Canadian pediatricians and family physi-
breastfeeding-support.html. Accessed July 2013.
cians. However, physicians with a breastfeeding interest 4. Leavitt G, Martinez S, Ortiz N, et al. Knowledge about breast-
would be expected to have better knowledge, confidence, feeding among a group of primary care physicians and resi-
beliefs, and attitudes than those with no or little breastfeeding dents in Puerto Rico. J Community Health. 2009;34(1):1-5.
interest, therefore leading to superior results. Also, resident 5. Finneran B, Murphy K. Breast is best for GPs—or is it?
participation depended on a multistep process (distribution of Breastfeeding attitudes and practice of general practitioners
letters by program directors, resident initiating communica- in the mid-west of Ireland. Ir Med J. 2004;97(9):268-270.
tion with our research assistant), likely causing the low par- 6. Nakar S, Peretz O, Hoffman R, et al. Attitudes and knowledge
ticipation level and resulting in only the most motivated on breastfeeding among paediatricians, family physicians,
participating in the study. Finally, although the survey instru- and gynaecologists in Israel. Acta Paediatr. 2007;96(6):
ment was piloted and reviewed, the questionnaire had not 848-851.
7. Al-Nassaj HH, Al-Ward NJA, Al-Awqati NA. Knowledge,
previously been used, the scale used has not been validated,
attitudes and sources of information on breastfeeding among
and psychometric testing was not performed. However, medical professionals in Baghdad. East Mediterr Health J.
Cronbach’s alpha values were found to be acceptable or 2004;10(6):871-878.
excellent for all domains of the questionnaires except beliefs. 8. Al-Zwaini EJ, Al-Haili SJ, Al-Alousi TM. Knowledge of
Iraqi primary health care physicians about breastfeeding. East
Mediterr Health J. 2008;14(2):381-388.
Conclusion 9. Brodribb W, Fallon A, Jackson C, et al. Breastfeeding and
The results of our study show that breastfeeding knowledge, Australian GP registrars—their knowledge and attitudes. J
confidence, beliefs, and attitudes are suboptimal among Hum Lact. 2008;24(4):422-430.
10. Freed GL, Clark SJ, Sorenson J, et al. National assessment of
Canadian primary care physicians. This study uncovered spe-
physicians’ breast-feeding knowledge, attitudes, training, and
cific areas of concern in physicians’ knowledge, confidence, experience. JAMA. 1995;273(6):472-476.
beliefs, and attitudes with regard to breastfeeding. Based on 11. Royal College of Physicians and Surgeons of Canada. Objectives
these results, targeted educational tools specifically address- of training pediatrics. http://www.macpeds.com/documents/
ing the areas identified through our survey can be developed Objectives_Training_Requirements_Pediatrics_000.pdf.
and implemented at the residency level. By optimizing the Published 2008. Accessed July 2013.
knowledge, confidence, beliefs, and attitudes of frontline 12. The College of Family Physicians of Canada. Defining compe-
physicians, we hope to increase breastfeeding exclusivity and tence for the purposes of certification by the College of Family
duration and improve the health of our children. Physicians of Canada: the evaluation objectives in family medi-
cine. Working Group on the Certification Process. http://www.
cfpc.ca/EvaluationObjectives/. Published 2010. Accessed July
Authors’ Note 2013.
The full questionnaire can be obtained from the main author at 13. Dillman DA. Mail and Internet Surveys: The Tailored Design
[email protected]. Method. 2nd ed. New York, NY: John Wiley & Sons, Inc; 2000.

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015

12 Journal of Human Lactation 

14. Registered Nurses’ Association of Ontario. Breastfeeding: 20. Pound CM, Gaboury I. The impact of jaundice in the new-
Fundamental Concepts—A Self-Learning Package. Toronto, born period on length of breastfeeding. Paediatr Child Health.
Canada: Registered Nurses’ Association of Ontario; 2006. 2009;14(7):445-449.
15. Feldman-Winter L, Barone L, Milcarek B, et al. Residency 21. Wood L, Wade K, Forham H, et al. Breastfeeding in Toronto—
curriculum improves breastfeeding care. Pediatrics. Promoting Supportive Environments. Technical Report.
2010;126(2):289-297. Toronto, Canada: Toronto Public Health; March 2010.
16. Handfield-Jones R, Rainsberry P. Certification examination 22. Howard CR, Howard FM, Lawrence RA, et al. Office prenatal
of the College of Family Physicians of Canada. Can Fam formula advertising and its effect on breast-feeding patterns.
Physician. 1996;42:957-969. Obstet Gynecol. 2000;95(2):296-303.
17. World Health Organization. The Optimal Duration of Exclusive 23. Kim HS. Attitudes and knowledge regarding breastfeeding: a
Breastfeeding: Report of an Expert Consultation. Geneva, survey of obstetrics residents in metropolitan areas of South
Switzerland: WHO Department of Nutrition for Health and Korea. South Med J. 1996;89(7):684-688.
Development, Department of Child and Adolescent Health and 24. Brodribb W, Fallon A, Jackson C, Hegney D. The relation-
Development; 2001. ship between personal breastfeeding experience and the
18. Dunn S, Davies B, McCleary L, et al. The relationship between breastfeeding attitudes, knowledge, confidence and effec-
vulnerability factors and breastfeeding outcome. J Obstet tiveness of GP registrars. Matern Child Nutr. 2008;4(4):
Gynecol Neonatal Nurs. 2006;35(1):87-97. 264-274.
19. Ahluwalia IB, Morrow B, Hsia J. Why do women stop
25. Asch S, Connor SE, Hamilton EG, et al. Problems in recruit-
breastfeeding? Findings from the Pregnancy Assessment and ing community-based physicians for health services research. J
Monitoring System. Pediatrics. 2005;116(6):1408-1412. Gen Intern Med. 2000;15(8):591-599.

Downloaded from jhl.sagepub.com at Gazi University on January 5, 2015