Quality of care, maternal attitude and common physician practices across the socio-economic spectrum: a community survey

Arch Gynecol Obstet (2010) 282:245–254 DOI 10.1007/s00404-009-1214-x

M A T ER N O - F E T A L M E D I C I N E

Quality of care, maternal attitude and common physician practices across the socio-economic spectrum: a community survey Rinku Sengupta Dhar · Jitender Nagpal · Vijaylakshmi Bhargava · Aarti Sachdeva · Abhishek Bhartia

Received: 28 March 2009 / Accepted: 13 August 2009 / Published online: 2 September 2009 © Springer-Verlag 2009

Abstract Purpose To evaluate the quality of maternity care, women’s perception of the quality of care and factors inXuencing the same in Delhi. Method Women who delivered a viable live birth in the past 6 months were recruited from South Delhi by a twostage stratiWed cluster randomized sampling. In stage 1, two colonies each from three predeWned economic strata (high-, middle- or low-income areas) were selected by simple-random sampling. In stage 2, a sequential house-to-house survey was conducted in each selected colony. The information was collected by interview and review of medical records. Results A total of 5,279 houses were screened to recruit 249 subjects. Several disparities were notable. 25.2 and 14.8% of the women from middle- and lower-income areas delivered by cesarean section, while the rate was 53.6% in the higher-income areas. In women from lower-income areas urine testing was not done in 76.9% and blood sugar was not tested in 18.2%, while in high-income areas 44.6%

R. S. Dhar · V. Bhargava Department of Gynecology and Obstetrics, Sitaram Bhartia Institute of Science and Research, B-16, Qutab Institutional Area, New Delhi 110016, India J. Nagpal Department of Pediatrics, Sitaram Bhartia Institute of Science and Research, B-16, Qutab Institutional Area, New Delhi 110016, India e-mail: [email protected] J. Nagpal · A. Sachdeva · A. Bhartia (&) Department of Clinical Epidemiology, Sitaram Bhartia Institute of Science and Research, B-16, Qutab Institutional Area, New Delhi 110016, India e-mail: [email protected]

had ¸10 antenatal visits and 87.8% had ¸3 ultrasounds. Of the women who experienced labor only 11% received support from a friend or family member, 4.4% received any medicine/measures for pain relief, 44.3% were allowed to walk and 14.6% were catheterized. Conclusion Health care providers are unable to meet national minimal care standards in poorer areas, while over investigation and over intervention appear prevalent in higher-income areas. This warrants a more representative evaluation to enable more equitable and evidence-based practice. Keywords

Quality of care · Maternity · Survey

Introduction “Quality of medical care” has been deWned in a broader sense as the capacity of the elements of that care to achieve legitimate medical and non-medical goals [1]. This has been interpreted to include the accessibility of services, the infrastructure facilities, the specialization and expertise of manpower, the ability of this manpower to follow guidelines and client satisfaction. From a narrower perspective, it has been deWned as “the degree to which health care services for individuals and populations increase the likelihood of desired health outcomes and are consistent with current professional knowledge” [2], while solely from the physician standpoint it implies consistency of practices with current professional knowledge. In the context of maternity care in India, suboptimal quality of maternity care, including inability to meet minimal care standards, lack of antenatal education, high intervention rates and non-compliance to evidence-based guidelines, may be a major public health concern because

123

246

Arch Gynecol Obstet (2010) 282:245–254

of its known contribution to high maternal mortality/morbidity [3], cost of care and adverse impact on service utilization. Several studies from western countries have documented the quality of maternity care in the earlier deWned context [4, 5]. This has enabled translational research into programs for improvement and monitoring of subsequent temporal changes [6]. In India, several large surveys like the National Family Health Survey (NFHS) [7] and the District Level Household Survey (DLHS) [8] have been attempted to provide insight into the quality of maternity care delivered at the population level. However, they have primarily focused on Reproductive and Child Health Program (RCH) [9] related outcomes like three antenatal visits, universal iron supplementation and immunization which do not reXect ideal practice guidelines but global minimal care objectives. Also, they do not provide any insight on the socio-economic variation in the quality of care, physician practice and maternal attitude or perception of the same. Thus, we proposed the current study to evaluate the quality of maternity care from Delhi in terms of the degree to which the physician practices are consistent with current professional knowledge, the factors inXuencing the same and women’s perception of the quality of care.

Materials and methods This survey was conducted in six selected colonies of South Delhi (one of the nine districts of Delhi) between February and April 2007. A summary of the study design is Fig. 1 Summary of study design

presented in Fig. 1. Women who delivered a viable live birth (after 28 weeks of gestation) in the past 6 months were included in the survey. The maximal acceptable duration after delivery was chosen on the basis of a prior trial run (n = 20) where women with greater time elapsed since delivery were very inaccurate in recall of postnatal events in comparison with available documentation. Women to whom the survey questionnaire could not be administered (unable to communicate due to serious illness or physical/ mental disability), women with major illnesses—cardiac, renal, hepatic, intestinal, neurological disease which is requiring continuing treatment or had required hospital admission for >1 week prior to recruitment (within the last 1 year) and women who had delivered outside Delhi were excluded. The survey intended to evaluate the quality of maternity care, common physician practices, maternal attitude and perception, the cost of maternity care (data published elsewhere [10]) and the postnatal quality of life (data published elsewhere [11]) in these women. The questionnaire was prevalidated on 20 postnatal patients visiting the outpatient department of our institute by video recording of qualitative and semi-structured interviews. Questionnaire administration by the two research team was subsequently standardized on another 20 postnatal patients admitted at our institute. Subjects in this survey were recruited from South Delhi (one of the nine districts constituting the state of Delhi) by a two-stage stratiWed cluster randomized sampling. In stage 1, two colonies (clusters; a colony is a small administrative unit in Delhi usually 1–2 km2) each from areas belonging to

ESTIMATED 4337* ELIGIBLE WOMEN-DELIVERED IN SOUTH DELHI DISTRICT IN LAST SIX MONTHS- 255 COLONIES; 70 IN HIG; 78 IN MIG; 107 IN LIG (BASED ON PROPERTY TAX CLASSIFICATION)

TWO COLONIES EACH BELONGING TO HIG (n=70), MIG (n=78), LIG (n=107) WERE SELECTED BY SIMPLE RANDOM SAMPLING

EACH OF THESE COLONIES WERE SEQUENTIALLY VISITED FOR ENROLLING MOTHERS FROM EACH COLONY

FIRST 50 MOTHERS FROM EACH SOCIOECONOMIC STRATUM • HOUSE TO HOUSE SURVEY • DETAILED QUESTIONNAIRE – MOTHER & CHILD CARE EDUCATOR AND STAFF NURSE • HEMOGLOBIN – HEMOCUE • BP, HEIGHT, WEIGHT, WAIST CIRCUMFERENCE

INCLUSION CRI Delivered in last 6 months Viable birth-born after 28 weeks Live birth- told by PCP EXCLUSION CRI Unable to communicate Major illness Delivered outside Delhi

* Estimated by the formula: n = (population of South Delhi × birth rate)/2,000

123

Arch Gynecol Obstet (2010) 282:245–254

socio-economic categories A or B [high-income group (HIG) 70 colonies; stratum 1], C or D [middle-income group (MIG) 78 colonies; stratum 2] and E, F or G [lowincome group (LIG) 107 colonies; stratum 3] [administrative classiWcation based on income used in Delhi for determining property tax; range A–G; ‘A’ highest] were chosen from South Delhi by simple random sampling. Details of the study design have been reported earlier. Deliberate over sampling (beyond the proportional representation) from the higher- and middle-income colonies was done to allow meaningful inferences on socio-economic inXuences. The sampling was restricted to South Delhi due to the convenience of proximity to our institute. However, despite the restricted geographical area, this socio-economically diverse population is served by a complete spectrum of governmental and private sector health services providing an ideal resource for this evaluation. In stage 2, a house-tohouse survey was conducted in one of the four random directions (north to south, south to north, east to west or west to east) in the selected colony proceeding in a sequential manner to screen for subjects fulWlling the selection criteria till a minimum of 50 subjects were recruited in each income category over an allotted period of 2–4 weeks. The subjects were then given a date and time for questionnaire administration and examination within 2 weeks of the initial visit. An informed consent was obtained from each subject. No incentives were given other than free test results of hemoglobin, blood pressure, weight and height measurements. The project was approved by the institutional ethics committee. The baseline information, recorded on a standard pretested questionnaire, included age, education, medical beneWts (government and private medical insurance or reimbursement), gross family income, place of antepartum, intrapartum and postpartum care and the duration of hospital stay. The questionnaire was designed to provide information on minimal care objectives from the national RCH program like iron supplementation, tetanus immunization, number of antenatal visits and trained attendance during delivery using relevant questions based on the NFHS 3 questionnaire [12]. Common physician practices, including potentially overused or misused investigations and interventions like ultrasound (USG), catheterization, IV Xuids, shaving of pubic hair etc., were also documented. Information on the knowledge and attitude of the mother toward certain interventions was collected using questions based on the ‘Listening to mothers’ survey [13]. The questionnaire was administered to each subject by a health counselor and a staV nurse (interviewer administered) who recorded the weight, height, waist circumference and blood pressure. The questionnaire was translated into Hindi and back translated into English to allow

247

administration in either language. Weight was recorded using manual spring dial weighing machine to the accuracy of 0.5 kg. Height was recorded using a ‘SECA®’ stadiometer with a minimum recording of 0.1 cm. Blood pressure was recorded using an ‘OMRON®’ electronic instrument accurate to 1 mmHg and was pretested on 20 subjects against a standard mercury barometer. Sample size considerations This study was intended to be a precursor to a proposed larger survey covering whole of Delhi and was partly intended to provide estimates of the design eVect and variance in expenditure to enable sample size calculations on this basis. In the context of the pilot survey and in the absence of substantive information on the subject from the area, it was presumed that to enable calculation of intracluster correlation coeYcient (or design eVect) and to provide reasonable insight on socio-economic trends a minimum of 50 subjects from each income category with two clusters (n = 25) each would be suYcient [14]. This was also anticipated to provide an estimate of logistic requirements over a period of 6–12 weeks (2–4 weeks in each income category). Data analysis Data entry and analysis was done using Epi-info2002 and SPSS v 13.0. The complex samples procedure of the SPSS was used to correct for the loss of sampling eYciency, resulting from the use of cluster sampling as opposed to simple random sampling. The procedure provided weighted estimates for the whole population on the basis of the results from individual clusters, proportional representation or weight of the particular cluster (the weight given to every case is equal to the inverse of the probability of selecting the individual which in turn would depend on the size of the cluster he/she is chosen from and the number of clusters) and the inter- and intra-cluster variation in the outcome variable.

Results A total of 5,279 houses (257–1,502 per cluster) were visited by the survey team, and 282 women meeting the inclusion criteria were identiWed. Six women who delivered outside Delhi were excluded from the analysis. Another 27 were excluded because they refused consent or were not available despite three follow-up visits. Thus, data from 249 women were analyzed. Of the 249 women, 50 women were from high-income areas, whereas 99 and 100 belonged to the middle- and low-income areas, respectively.

123

248

The socio-demographic proWle of the subjects is presented in Table 1. The mean age of women was 26.9 years, 44.9% women were primiparous and 36.8% were anemic at the time of survey. One-third of the women from the lowerincome areas had primary school education or lower. Of the 249 women, 54.9% women belong to families with less than Rs. 5,875 monthly family income, 73.9% women had never worked and only 17.9% had some medical beneWts. Majority of women’s (79%) had hospital delivery, 12.8% had delivery at small institutions which included facilities with less than 25 beds and only 8.2% delivered at home. Cesarean section rate was recorded to be 30.4%, and the average baby weight at birth was 2.9 kg. The quality of antenatal, delivery and postnatal care services is summarized in Table 2. Approximately 10% LIG mothers had less than three antenatal visits, while 44.6% HIG mothers had more than ten antenatal visits. 91.5% of the women received iron supplementation, while 92.1% were fully immunized against tetanus. During antenatal period, 49% mothers had more than three ultrasounds done, and the number of ultrasounds increased with the income group (Fig. 2). In the study, only 41.2% HIG mothers were aware of labor analgesia. Of the total, 66.5% mothers initiated breast-feeding within 24 h, and 42.9% mothers were allowed to touch the babies within a few minutes. After delivery, only 49.5% had hospital visit for regular checkup, and 35.3% mothers were advised about family planning. The same data were then stratiWed by the type of institution involved in the care (government, private and nursing home), education and employment status. It was noted that 90–95% women received two or more doses of tetanus vaccination [Government 94.6% (92.4–96.2), Private 88.8 (81.0–93.6), Nursing home 93.5% (82.6–97.6%)] for other aspects of care, including urine test, blood sugar test, blood pressure testing, iron supplementation and hemoglobin estimation non-institutional/home deliveries a lower population of women received the standard of care. In the case of ultrasounds 61.1% of women delivering in private hospitals had more than three ultrasounds compared with 6.2% for non-institutional and 10.7% or government hospitals. Results for stratiWcation by education and employment were similar to those for income (data not presented). Table 3 highlights the knowledge, attitude and awareness of women and practices of their care providers. It is noteworthy that »79% of the LIG mothers reported that their pregnancy was planned. Merely 11% mothers had some form of labor support provided by a friend or family members. As depicted, 28.2% mothers reported some form of labor induction with lower rates in LIG (»19%) mothers. 66.2% mothers delivering vaginally were given an episiotomy. In 53.8% mother’s pubic hair was shaved, and 50% mothers were given enema. Less than half of the

123

Arch Gynecol Obstet (2010) 282:245–254

women were allowed to walk, eat or drink during labor. Only 6.6% mothers preferred cesarean section during the antenatal period, and 23% of mothers requested shifting their baby to the nursery for non-medical reasons. Onethird of mothers discard the colostrum portion of breast milk, concerned that it is deleterious to the baby.

Discussion As presented earlier term ‘quality of care’ in our survey and for most diseases and conditions refers to the status of the current practices in comparison with the evidence-based recommendations available from the expert bodies [15, 16]. However, in the context of maternity care in India, quality has invariably been presented in comparison with national or international minimal public health objectives which may be substantially short of evidence-based recommendations. The data from our survey, though preliminary, documents that substantial sections of the lower-income populace fail to meet some of the minimal national public health goals like universal iron supplementation and a minimum of three antenatal visits, while at the same time some interventions such as USG and phototherapy are being overused in the HIG. Also several widely prevalent practices like episiotomy, catheterization and shaving of pubic hair do not conform to the current evidence-based guidelines (routine episiotomy, urinary catheterization and shaving of pubic hair is not recommended) [17]. Institutional delivery rates, cesarean rates and proportion of the population receiving iron supplementation and tetanus vaccination were higher in the current survey compared to earlier reports from Delhi [12]. This may represent a geographical variation (the survey was restricted to one district) or may reXect temporal improvement. In the context of quality of maternity care, the National Population Policy adopted by the Government of India in 2000 [18] lays down socio-demographic goals to reduce maternal mortality ratio to 100 per 100,000, to ensure that 80% of the deliveries take place in institutions and to ensure that 100% of the deliveries are attended by trained personnel. Similarly, the RCH Program [9] recommends that as part of antenatal care women should receive two doses of tetanus toxoid vaccine (TT), adequate amounts of iron and folic acid (IFA) tablets or syrups to prevent and treat anemia and have at least three antenatal check-ups that include blood pressure checks and other procedures to detect pregnancy complications. Several government-sponsored and independent attempts have been made to assess the quality of care with reference to the stated public health goals, including the NFHS [7] and the DLHS [8] surveys. The NFHS-2 (2000) results show that more than one out of every three pregnant

Arch Gynecol Obstet (2010) 282:245–254

249

Table 1 Socio-demographic proWle of the population Characteristics

Overall (n = 249)

Age (years)a

According to area category HIG (n = 50)

MIG (n = 99)

LIG (n = 100)

26.9 (25.4–28.3)

29.9 (28.7–31.1)

26.2 (23.7–28.6)

24.8 (24.2–25.4)

1.9 (1.8–2.0)

1.6 (1.4–1.8)

2.1 (1.9–2.3)

2.0 (1.9–2.0)

Obstetric Historya Parity (n) Primi (%)

44.9 (38.0–51.9)

49.7 (32.7–66.8)

46.9 (40.3–53.7)

40.0 (37.6–42.4)

Abortions (n)

0.2 (0.1–0.3)

0.2 (0.0–0.5)

0.3 (0.1–0.5)

0.2 (0.1–0.2)

BMI (kg/m2)a

24.4 (22.4–26.4)

27.7 (24.9–30.5)

24.3 (22.4–26.2)

21.9 (20.9–22.9)

Anemic patients (%)a,b

36.8 (24.2–49.4)

20.9 (8.0–44.7)

53.2 (49.5–56.9)

75.0 (63.6–83.8)

18.0 (7.7–36.6)

0 (0)

17.5 (12.8–23.6)

32.3 (18.7–49.8)

Education level (%)a Illiterate/primary school Middle or high school

36.5 (26.5–47.9)

7.0 (0.8–42.3)

55.3 (15.9–89.0)

50.2 (32.2–68.2)

¸College education

45.5 (29.9–62.0)

93.0 (57.7–99.2)

27.2 (4.4–75.1)

17.4 (14.2–21.2)

Gross monthly family income (Rs.)a,c 50,000

27.1 (13.9–46.1)

77.4 (35.9–95.4)

2.4 (0.2–22.1)

0 (0)

Never worked

73.9 (59.4–84.6)

39.4 (34.2–44.9)

85.0 (67.3–93.9)

95.5 (90.3–98.0)

Working full time

6.6 (1.7–22.4)

17.8 (3.9–53.4)

2.4 (0.2–22.1)

0 (0) 0.6 (0.0–25.2)

Current employment statusa (%)

Working part time

5.8 (2.2–14.1)

14.5 (7.4–26.6)

2.5 (0.1–38.7)

Not working at present

13.6 (6.7–25.7)

28.3 (13.8–49.2)

10.1 (2.2–36.7)

3.9 (2.8–5.4)

Who had medical beneWtsd

17.6 (6.3–40.5)

34.3 (20.5–51.5)

23.9 (1.8–84.4)

1.3 (0.0–40.6)

Place of delivery Hospitale

79.0 (52.7–92.7)

83.5 (62.4–93.9)

83.2 (76.3–88.5)

73.4 (28.1–95.1)

Government

40.7 (33.2–48.6)

2.8 (0.2–35.8)

65.4 (45.7–80.9)

57.9 (25.0–85.1)

38.3 (25.5–53.1)

80.6 (65.8–90.0)

17.9 (4.1–52.8)

15.5 (11.9–19.9)

Private f

Non-institutional

12.8 (7.3–21.5)

16.5 (6.1–37.6)

14.3 (11.9–17.0)

9.1 (1.7–36.0)

Home

8.2 (1.1–40.8)

0 (0)

2.5 (0.1–38.7)

17.5 (3.9–52.8)

Mode of delivery CS

30.4 (17.3–47.7)

53.6 (41.5–65.3)

25.2 (13.7–41.9)

14.8 (4.9–36.7)

Elective CS

17.9 (8.4–34.2)

36.2 (21.9–53.5)

12.3 (4.4–29.7)

8.4 (4.8–14.1)

Emergency CS

12.5 (6.3–23.3)

17.4 (7.9–34.1)

13.0 (1.5–59.9)

6.4 (1.1–29.6)

NVD with perineum intact

21.6 (8.7–44.5)

1.4 (0.1–19.1)

15.5 (8.6–26.5)

40.7 (24.1–59.8)

NVD with epi

43.7 (38.0–49.5)

43.6 (37.3–50.1)

52.3 (35.4–68.7)

39.3 (34.4–44.5)

NVD with tear

1.3 (0.2–7.7)

0 (0)

3.3 (1.2–8.6)

1.3 (0.0–40.6)

Instrumental

3.0 (0.8–10.7)

1.4 (0.1–19.1)

3.6 (0.3–30.6)

3.9 (0.8–17.4)

Birth weight of baby

2.9 (2.8–3.0)

3.0 (3.0–3.1)

2.9 (2.6–3.2)

2.8 (2.6–2.9)

Data is presented as cluster adjusted mean (95% CI) or percentage (95% CI) taking into account the survey design. Results with statistically signiWcant diVerences are highlighted in bold  There was a statistically signiWcant diVerence on the HIG versus MIG and the HIG versus LIG comparisons  There was a statistically signiWcant diVerence on all three comparisons: HIG versus MIG, MIG versus LIG and HIG versus LIG a These items reXects the status of the women at the time of conducting the survey b Anemia was deWned as Hb · 11 g% c Gross family monthly income (posttax take-home family income) d Any OPD or IPD medical reimbursement e Hospital was deWned as >25 beds setup f Non-institutional delivery includes nursing home, private dispensary, government dispensary and individual practitioner home (clinic)

123

250

Arch Gynecol Obstet (2010) 282:245–254

Table 2 Quality of maternity care and common physician practices Overall (n = 249)

According to area category HIG (n = 50)

MIG (n = 99)

LIG (n = 100)

Antenatal care services Antenatal visits women who had ·3 (%)

4.7 (1.9–10.8)

0 (0)

0.8 (0.0–16.7)

10.3 (7.9–13.4)

Period of gestation at Wrst checkup (weeks) (%)

10.3 (9.1–11.2)

5.7 (5.0–6.5)

10.8 (8.6–13.1)

13.6 (11.6–15.5)

Women who took iron supplementation (%)

91.5 (83.3–95.8)

97.2 (67.3–99.8)

97.0 (93.4–98.7)

84.2 (80.0–87.7)

Tetanus toxoid 2 doses (%)

92.1 (89.8–93.9)

88.3 (83.4–91.8)

92.7 (87.7–95.7)

94.8 (93.0–96.2)

1 dose (%)

7.2 (5.6–9.1)

11.7 (8.2–16.6)

4.1 (3.1–5.3)

5.2 (3.8–7.0)

Not received (%)

0.7 (0.2–2.9)

0 (0)

3.3 (1.2–8.6)

0 (0)

Urine tested ¸2 visits (%)

53.9 (29.0–77.0)

90.2 (62.4–98.1)

57.8 (26.9–83.6)

23.1 (9.1–47.4)

Hemoglobin tested once or more (%)

94.8 (71.3–99.3)

100.0 (100.0–100.0)

97.1 (93.4–98.8)

89.6 (59.9–98.0)

Ultrasound Done (%)

93.2 (72.3–98.6)

100.0 (100.0–100.0)

97.1 (93.4–98.8)

85.7 (63.0–95.5)

Number (mean)

2.8 (1.9–3.7)

4.4 (3.7–5.2)

2.5 (1.9–3.0)

1.7 (1.1–2.3)

¸3 USG done (%)

49.1 (30.5–67.9)

87.8 (54.2–97.8)

41.2 (30.6–52.6)

22.5 (14.0–34.2)

Explained signs and symptoms of preterm labor (%)

21.1 (8.1–44.6)

44.5 (23.3–67.9)

7.1 (5.2–9.6)

9.6 (0.7–60.9)

Told about labor analgesia/pain relief (%)

14.9 (8.3–25.2)

41.2 (29.8–53.7)

3.3 (1.2–8.5)

0 (0–0)

Diet counseling givena (%)

24.8 (18.3–32.6)

38.9 (20.0–61.9)

22.4 (19.0–26.3)

14.9 (5.9–32.8)

20.0 (8.9–39.1)

44.2 (20.5–70.8)

14.5 (3.8–42.6)

6.7 (1.9–21.1)

Delivery care services Type of monitoring (continuous %) (n = 176)b Medications/measures for pain relief (%)

4.4 (1.4–13.0)

5.6 (0.4–46.0)

8.6 (6.3–11.6)

1.4 (0.3–7.3)

Who did not want pain relief in labor (%)

91.2 (82.0–96.0)

89.9 (52.6–98.6)

89.9 (87.6–91.8)

92.8 (87.7–95.9)

Delivered by traditional birth attendant (TBA/Dai) (%)

7.6 (1.1–38.0)

0 (0)

2.5 (0.1–38.7)

16.2 (3.7–49.4)

Availability of CS facility at the place of delivery (%)

86.7 (78.8–92.0)

90.2 (62.4–98.1)

90.6 (75.7–96.7)

81.9 (79.9–83.8)

Baby weighed at birth (%)

92.0 (58.7–98.9)

100.0 (100.0–100.0)

98.3 (71.0–99.9)

82.5 (47.2–96.1)

Shown baby within 0.5 h after birth (%)

85.8 (80.5–89.8)

90.6 (71.4–97.4)

83.3 (73.9–89.8)

83.2 (80.7–85.5)

Allowed to touch the baby within few minutes (%)

42.9 (35.8–50.3)

55.3 (40.8–69.0)

21.7 (10.1–40.8)

44.8 (42.1–47.6)

Hospital staV helped in initiation of breast feeding (yes) (n = 232)c (%)

69.6 (59.2–78.3)

96.2 (85.1–99.1)

57.4 (43.6–70.1)

51.5 (45.4–57.7)

Initiation of breast-feeding in