Abstract

Objective. To assess the quality of antenatal care with respect to anaemia, a common health problem, in a developing country.

Design. Cross-sectional study.

Setting. Rufiji rural district, coastal Tanzania.

Study participants. Three hundred and seventy-nine consecutively enrolled pregnant women from 16 randomly selected antenatal clinics, including 10 dispensaries, four rural health centres and two hospitals.

Methods. We noted availability of infrastructure for management of anaemia, observed provider–client interaction, and interviewed women who attended antenatal clinics. An observer and health workers measured haemoglobin levels and their measurements were compared.

Main outcome measures. Quality of antenatal care, anaemia in pregnancy, and maternal satisfaction.

Results. Half of the rural health clinics had no instrument to measure haemoglobin. The majority (58%) of pregnant women were not checked for anaemia at all, 10% were clinically examined, and 37% had haemoglobin assessed. The agreement between health workers’ (using Tallqvist) and the observers’ (using HemoCue) measurements of haemoglobin was poor to fair. The prevalence of anaemia (Hb <10.5 g/dl) and severe anaemia (<7.0 g/dl) was 58% and 6.2%, respectively, but overall only 4% of the anaemic pregnant women had specific action taken within the antenatal care program.

Conclusion. Deficiencies in quality of screening, diagnostic information, and individual counselling need to be addressed before any impact of the antenatal care programme on anaemia can be expected.

Antenatal care is one of the pillars of Safe Motherhood interventions that is believed to reduce maternal and perinatal mortality if carried out properly [1]. Universal access to antenatal care is a matter of priority in both developed and developing countries [2]. However, detailed information about the actual quality and effectiveness in practice is less often available or searched for [3,4]. In theory, proper assessment of risk factors, and a series of health examinations with pre-defined content should enable health personnel to uncover conditions in the mother that may threaten her and or her foetus during pregnancy. The condition may then be treated or monitored to secure a better outcome [5].

In Tanzania, 97% of pregnant women attend antenatal care (ANC), and 70% do so at least four times [6]. In spite of this, the prevalence of anaemia, using a cutoff point of haemoglobin (Hb) <10.5 g/dl, was 60% in antenatal clinics in Dar es Salaam in 1990–92 [7]. Maternal mortality in Tanzania from the 1996 Demographic and Health survey is estimated at 529 per 100 000 live births [4], but others have estimated it to be 770 per 100 000 live births [8]. Anaemia is ranked as the fourth commonest cause of maternal mortality after abortion, haemorrhage and hypertension/eclampsia, contributing to 11% of maternal deaths [9]. The disparity between high antenatal care attendance, high prevalence of anaemia in pregnancy, and high maternal mortality raise questions about the quality of care provided, and particularly the management of anaemia in pregnancy at antenatal clinics.

We investigated the quality of antenatal care with respect to screening for and management of anaemia, and the relation between performance and recommended routines in antenatal clinics. Specifically, the study assessed structural, process, and maternal satisfaction components of quality of management of anaemia at antenatal clinics in selected health care institutions in the Rufiji district, Tanzania.

Subjects and methods

Design

A cross-sectional study design was used.

Setting

Rufiji district is one of the six districts in the coastal region of Tanzania. In 1998 the district had a population of 180 000, based on the 1988 Census, and an annual population growth of 2.3%. Geographically, the Rufiji river intersects the district from west to east, dividing it into flood plain, coastal-delta, and plateau (hill) zones. The road network in the district is not passable during the rainy season, and the limited transport system makes people dependent on the available health services within the district.

ANC is provided at two hospitals, four rural health centres (RHCs), and 48 dispensaries. The health workers provide ANC for low risk women at dispensaries and RHCs, and refer women with risk factors to a hospital according to guidelines stipulated on the antenatal card. Pregnant mothers at all levels of health care are supposed to receive iron and folate supplements free of cost at each visit.

Study population and sampling

The study population included pregnant women who attended antenatal clinics and was stratified by type of unit providing ANC. In two strata, hospitals and RHCs, all units were included in the sample, while in the third stratum 10 out of 48 dispensaries were randomly selected. Two of the selected dispensaries were owned by non-government institutions. This fulfils recommendations to cover at least 25% of the health care institutions in the area when assessing quality of care [10,11]. At the hospitals and RHCs the first 30 women, and at dispensaries the first 20 women of 1 day were invited to participate in the study; in total, 380 women were recruited.

Data collection

The research team consisted of one researcher (D.P.U.) and two qualified nurse midwives (research assistants) who visited each facility for 2 days. An introductory meeting was held on the first day in each facility to establish contact with staff and to explain the aim of the study. The research team was introduced as not coming from the Ministry of Health, and it was emphasized that the findings of this study would by no means be used against any staff member. On the second day of the visit, women attending ANC were observed and interviewed after obtaining informed consent.

A trained research assistant was stationed in each consultation room to observe the interaction between the health worker and the pregnant woman. Using an observation guideline, the observer took notes on the qualifications of the staff member, whether clinical examination for anaemia was performed (see Definitions), and whether an investigation into haemoglobin was requested. She also recorded whether feedback was given on risk factors detected, and the type of counselling given or action taken.

Haemoglobin assessment

To estimate haemoglobin levels, the health workers used Tallqvist, the only simple subjective colorimetric method available in most of these institutions. A finger-prick drop of blood was placed on a strip of absorbent paper and was left to dry. The blood colour was compared to the standard Tallqvist color scale, which is calibrated at an interval of 10 units (Figure 1). The observer used the HemoCue haemoglobinometer machine to estimate haemoglobin. After calibration of the machine using a control cuvette provided by manufacturer, the standard cuvette was filled with a drop of blood from a finger prick and haemoglobin values were read and recorded to one decimal point. In comparing the two scales, a value of 14.8 g/dl in the HemoCue scale falls within a point interval of 100% in the Tallqvist scale. Similarly 11.0 g/dl falls next to 80%, 10.5 g/dl to 70%, 8.5 g/dl to 60%, and 7.0 g/dl to 50%, as illustrated in Figure 1. Many health workers recorded an average figure (e.g. 55%) of the blood colour that falls between two figures (e.g. 50% and 60%) in the corresponding Tallqvist color scale.

Figure 1

Comparative scale of Tallqvist and HemoCue values for haemoglobin estimation.

Figure 1

Comparative scale of Tallqvist and HemoCue values for haemoglobin estimation.

After each consultation, the observer checked the haemoglobin using the HemoCue haemoglobinometer. For ethical reasons mothers found by the observer to be anaemic (Hb ≤8.5 g/dl) were returned to the health worker with their results for further investigations, treatment, or referral. We did not register what kind of action was taken in these cases. To validate the observations, exit interviews of pregnant mothers were performed to assess their experience from the care they received. Questions were asked on the type of investigations or examinations done, feedback on detected problems, individual counselling or medication, and satisfaction with the service received. During exit interviews haemoglobin values reported by health workers were obtained from the ANC card. However, even though clinical examination for pallor was done, this was not always recorded. Therefore the health worker’s assessment of the mother as anaemic could only be obtained from the mothers who had received counselling.

The study used explicit measurements of components of ANC, relevant to management of anaemia in pregnancy. In each study site, an observation checklist was used to assess the availability of facilities for detection and management of anaemia. These included the presence of qualified staff and equipment for the estimation of haemoglobin. A standard questionnaire with closed- and open-ended questions in Kiswahili was used during exit interviews. All research assistants were trained in the use of the HemoCue for estimation of haemoglobin.

Definitions

Definitions of the terms used in this study were as follows.

Trained staff: those who have been exposed to midwifery in their basic training. Maternal and child health aide (MCH aide) with a 2 year training period was considered the lowest cadre of trained staff.

Clinical examination for anaemia: when health workers checked for pallor, at least in the conjunctiva.

Laboratory investigation for haemoglobin: having methods like Tallqvist, calorimetric and so forth for measuring haemoglobin.

Adequate structure: (1) the availability of trained staff; (2) the presence of method for investigating haemoglobin; (3) iron/folate tablets; and (4) anti-malarial tablets.

Adequate anaemia management: when a pregnant woman (1) was attended by a trained staff member; (2) was clinically examined for anaemia; (3) had a haemoglobin investigation performed; (4) was given results of her examination/investigation; and (5) was counselled according to the results.

These components were considered to be the standard criteria against which quality of antenatal care was assessed. The operational Hb cutoff level of ≤8.5 g/dl was used for moderate to severe anaemia that needed further investigation, treatment or referral according to guidelines in the antenatal card.

Statistical analysis

The collected information was coded, computerised, checked for consistency and analysed using Epi Info version 6, STATA version 6, and Excel 97 statistical and graphic presentation software. Diagnostic agreement between methods was assessed by the Kappa coefficient, with a K-value ≤0.20 indicating poor agreement, 0.21–0.40 fair agreement, 0.41–0.60 moderate agreement, 0.61–0.80 substantial agreement, and ≥0.81 almost perfect agreement between assessors [12]. ANOVA (analysis of variance) was used to test whether the difference in proportions of mothers assessed in health care institutions was due to a random variation or not.

Results

Three hundred and seventy-nine women were enrolled in the study, and 370 (98%) had their haemoglobin investigated by the observer. Nine women could not be traced by the observer for haemoglobin assessment after the health workers had examined them. Using WHO criteria of Hb <11.0 g/dl for anaemia, 69% of women were anaemic (Table 1). The median haemoglobin was 9.8 g/dl, with a range of 4.4–15.3 g/dl. The prevalence of anaemia differed significantly according to type of health facility, being less common in the hospital, with levels of <11.0 g/dl (P = 0.0002) and <10.5 g/dl (P = 0.015). The differences in prevalence of severe anaemia according to type of health facility were not significant.

Table 1

Prevalence of anaemia in pregnancy by health care institution for different cutoff points, as assessed by an observer with HemoCue

Haemoglobin Health care institution Total P value 
level (g/dl) Dispensary RHC Hospital No. (ANOVA) 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 

 
<11.0 123 66.5 100 80.6 32 52.5 255 68.0 0.0002 
<10.5 104 56.2  82 66.1 27 44.3 213 57.6 0.015 
 ≤8.5  49 26.5  36 29.0 15 24.6 100 27.0 0.810 
 <7.0   9  4.9   9  7.3  5  8.2  23  6.2 0.544 
No. of women 185  124  61  370   
Haemoglobin Health care institution Total P value 
level (g/dl) Dispensary RHC Hospital No. (ANOVA) 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 

 
<11.0 123 66.5 100 80.6 32 52.5 255 68.0 0.0002 
<10.5 104 56.2  82 66.1 27 44.3 213 57.6 0.015 
 ≤8.5  49 26.5  36 29.0 15 24.6 100 27.0 0.810 
 <7.0   9  4.9   9  7.3  5  8.2  23  6.2 0.544 
No. of women 185  124  61  370   

RHC, rural health centre.

Adequacy of structural resources according to the requirements in health care institutions is presented in Table 2. None of the government dispensaries, but both non-governmental dispensaries, had Tallqvist for assessment of haemoglobin, whereas all RHCs and both hospitals could use this method. Both hospitals had colorimeters, but they were not functioning. Iron/folate tablets were available in all health care institutions except in one dispensary, and two dispensaries had no chloroquine tablets available for malaria chemoprophylaxis, which was a drug of choice at the time of data collection. Only the hospitals could perform investigations examining blood for the presence of malarial parasites and examining stools for the presence of helminths. On average, a supervision visit to each health facility by the district health management team or the MCH coordinator was made every 3–4 months.

Table 2

Number of health care institutions with adequate structural facilities for screening, management, and prevention of anaemia during pregnancy

Presence of: Health care institution
 
Total
 
 Dispensary RHC Hospital No. 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 
Trained staff  9  90 100 100 15  93.8 
Haemoglobin instrument1  2  20 100 100  8  50.0 
Iron/folate tablets  9  90 100 100 15  93.8 
Anti-malarial tablets  8  80 100 100 14  87.5 
Number of institutions 10 100 100 100 16 100 
Presence of: Health care institution
 
Total
 
 Dispensary RHC Hospital No. 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 
Trained staff  9  90 100 100 15  93.8 
Haemoglobin instrument1  2  20 100 100  8  50.0 
Iron/folate tablets  9  90 100 100 15  93.8 
Anti-malarial tablets  8  80 100 100 14  87.5 
Number of institutions 10 100 100 100 16 100 

RHC, rural health centre.

1The Tallqvist colour scale in all cases.

Out of 370 women studied, 41% were screened for anaemia by a trained health worker using either clinical examination or the Tallqvist method (Table 3). There was a significant difference between health care institutions in the proportion of women assessed for anaemia (P < 0.01). Few women were clinically examined, even at dispensaries, where this was the only feasible option.

Table 3

Percentage of pregnant mothers assessed for anaemia clinically and using the Tallqvist method by health care institution

Method for assessment Health care institution
 
Total
 
P value 
 Dispensary RHC Hospital No. ANOVA 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 

 
Clinically  27 14.5   5  4.0  4  6.6  36  9.7 0.0089 
Tallqvist  12  6.4  87 70.1 37 60.6 136 36.7 0.0001 
At least one method  29 15.6  87 70.1 38 62.2 154 41.6 0.0001 
No. of women 185  124  61  370   
Method for assessment Health care institution
 
Total
 
P value 
 Dispensary RHC Hospital No. ANOVA 
 No.
 
%
 
No.
 
%
 
No.
 
%
 

 

 

 
Clinically  27 14.5   5  4.0  4  6.6  36  9.7 0.0089 
Tallqvist  12  6.4  87 70.1 37 60.6 136 36.7 0.0001 
At least one method  29 15.6  87 70.1 38 62.2 154 41.6 0.0001 
No. of women 185  124  61  370   

RHC, rural health centre.

Figure 2 shows the distribution of anaemia at different cutoff points in a subsample of 136 women who were investigated by both an observer and the health workers. Values around the cutoff points recorded by health workers were higher than those of the observers. The threshold for referral in these clinics is ≤60%, which corresponds approximately to Hb ≤8.5 g/dl.

Figure 2

Distribution of haemoglobin values in women assessed with the Tallqvist method by health workers.

Figure 2

Distribution of haemoglobin values in women assessed with the Tallqvist method by health workers.

Tallqvist values corresponded to a wide range of HemoCue values (Figure 3). The scattergram showed fewer (n = 14) cases below the <60% line according to Tallqvist than (n = 24) cases below the Hb 8.5 g/dl line according to HemoCue.

Figure 3

Scatter graph for haemoglobin measurements with HemoCue by the observer and with Tallqvist by the health worker. Lines indicate the cutoff points of <60% and <8.5 g/dl, respectively, which according to the guideline should indicate a need for referral.

Figure 3

Scatter graph for haemoglobin measurements with HemoCue by the observer and with Tallqvist by the health worker. Lines indicate the cutoff points of <60% and <8.5 g/dl, respectively, which according to the guideline should indicate a need for referral.

The degree of agreement in diagnosis of anaemia in 136 women as assessed using the Tallqvist or the HemoCue methods was calculated using the Kappa index (Table 4). HemoCue cutoff points for anaemia were Hb <11.0, <10.5, and ≤8.5 g/dl, and corresponding Tallqvist values were 75%, 70%, 60%, and 50%. There was fair agreement in mild and moderate levels of anaemia (K = 0.43 and 0.41 at Hb <11.0 and 10.5 g/dl, respectively), but very poor agreement for the most anaemic women with Hb ≤8.5 and <7.0 g/dl (K = −0.18 and −0.01, respectively).

Table 4

Agreement in diagnosis of anaemia between observers (HemoCue) and health workers (Tallqvist) at different cutoff points in 136 women investigated with both methods.

Haemoglobin measurements
 
Kappa coefficient 95% confidence interval 
HemoCue
 
Tallqvist
 
  
(g/dl) Number
 
(%)
 
Number
 
  
<11.0 86  75 111  0.41  0.23–0.59 
<10.5 66  70  95  0.43  0.27–0.59 
 ≤8.5 30 ≤60  60 −0.18 −0.31–(−0.04) 
 <8.5 24  60  14 −0.12 −0.28–0.03 
 <7.0 10  50   1 −0.01 −0.30–0.27 
Haemoglobin measurements
 
Kappa coefficient 95% confidence interval 
HemoCue
 
Tallqvist
 
  
(g/dl) Number
 
(%)
 
Number
 
  
<11.0 86  75 111  0.41  0.23–0.59 
<10.5 66  70  95  0.43  0.27–0.59 
 ≤8.5 30 ≤60  60 −0.18 −0.31–(−0.04) 
 <8.5 24  60  14 −0.12 −0.28–0.03 
 <7.0 10  50   1 −0.01 −0.30–0.27 

The proportion of anaemic mothers who received appropriate counselling according to severity of anaemia is presented in Table 5. Only 8–10% of the women found to be anaemic during routine examination had any action taken by health workers, as did only 3–4% of those with documented anaemia in the whole study group. No difference was found related to severity of anaemia. About 96% of all women attending the ANC clinics received routine chemoprophylactic iron and folate tablets, mostly by an untrained assistant and without any relation to the clinical findings. About 95% of the women reported that they were highly to moderately satisfied with the service. Of those diagnosed as having anaemia (Hb <11.0 g/dl), 97% were highly to moderately satisfied compared with 92% of the non-anaemic women. The question was not related to any specific aspect of health care.

Table 5

Number of anaemic women and percentage that received appropriate counselling grouped by haemoglobin level as assessed by an observer and by health workers

Haemoglobin (g/dl) Assessment by observer
 
Assessed by health workers
 
 No. of anaemic % received No. of anaemic % received 
 women
 
counselling
 
women
 
counselling
 
<11.0 255 3.1  96  8.3 
<10.5 215 3.8  74 10.8 
 ≤8.5 100 4.0  35 11.8 
 <7.0  23 4.3  10 10.0 
No. of women 370 157  8 
Haemoglobin (g/dl) Assessment by observer
 
Assessed by health workers
 
 No. of anaemic % received No. of anaemic % received 
 women
 
counselling
 
women
 
counselling
 
<11.0 255 3.1  96  8.3 
<10.5 215 3.8  74 10.8 
 ≤8.5 100 4.0  35 11.8 
 <7.0  23 4.3  10 10.0 
No. of women 370 157  8 

Discussion

This study has revealed a very high prevalence of anaemia in pregnant women in the Rufiji district of Tanzania, inadequate screening instruments, inadequate detection of anaemia by health workers using clinical and laboratory methods, and poor counselling of pregnant women who are anaemic, despite the fact that they generally reported satisfaction with the care.

The high prevalence of anaemia in this study is comparable to reports from many other parts of Tanzania [13,14] and tropical Africa, where prevalence rates of 50–60% during pregnancy have been reported [15,16]. Contributing factors to anaemia include inadequate iron stores due to multiple or closely spaced pregnancies, increased nutritional demand during pregnancy, poor food intake, hookworm infestation, malaria, and recurrent and chronic infections [17]. The strategies that have been recommended are based on major known causes of anaemia in the region, and include malaria and hookworm prophylactics and their treatment, as well as improved nutrition [18,19]. However, the success of these strategies depends on the ability of health services to provide the expected quality of care.

The first prerequisite for appropriate management of anaemia in antenatal care is available methods to identify the women with severe anaemia. None of the government dispensaries could perform even a simple assessment of haemoglobin, and this is a clear indicator of the inadequate funding of activities in the public health sector. In addition to this, not even clinical assessment was performed in 90% of the cases in these clinics.

It is also evident from this assessment of process quality that the Tallqvist method, which is very commonly used when laboratory resources are poor, seriously underestimates the most severe cases of anaemia, and that also for moderate anaemia the accuracy is less than acceptable. Thus, for the majority of anaemic pregnant women, their health problems would not even be detected.

A colour scale for assessing haemoglobin recently developed by WHO has been tested elsewhere and is recommended for use in screening for anaemia at antenatal clinics in settings where resources are limited [20]. Improvement in detemination of haemoglobin levels is essential in order to increase awareness of the problem and the motivation to look for the solution.

Although this cross-sectional study showed that most women received chemoprophylactic iron and folate tablets, another longitudinal study from Tanzania has reported frequent shortage of iron and folate tabs in antenatal clinics in similar settings [13]. Theoretically, moderate and mild iron deficiency may be corrected by oral therapy in women attending antenatal clinics mid-trimester, but studies have demonstrated variations in improvement of haemoglobin [3]. It is not known whether this is due to variation in absorption related to dietary factors, or to non-compliance of treatment due to side effects, or failure of health services to deliver treatment or to motivate women [3]. A study of compliance with iron supplementation during pregnancy in Tanzania showed a mean adherence to conventional iron tablets of 42%, but only 30% in those who had side effects [21]. Even if iron supplements are provided, women must be motivated to take them.

The fact that few anaemic women were screened by health workers, and that only 4% of those who were severely anaemic received appropriate individual counselling, casts doubt on the benefit of the current antenatal care programme in Tanzania for management of anaemia in pregnancy. Screening can only be useful if the results of any investigation are interpreted correctly and translated into an appropriate action.

Our paper did not address the issue of health workers’ attitudes toward the service they provide. A study conducted in South Africa reported that after successful completion of a Perinatal Education Programme (PEP), midwives in an obstetric unit could improve in documenting obstetric history. However, there was still no satisfactory detection of the obstetric problems and appropriate action was only taken in <12% of the cases [22]. Health education on quality improvement is an important need; however, more research is also recommended on factors related to poor performance of well trained staff in spite of availability of equipment and supplies, as was found in health centres and hospitals in this study.

The fact that the majority of women attending these clinics were highly to moderately satisfied with the care they received may indicate limitations of this study with respect to reliability and validity of questions on client satisfaction. Other qualitative research has shown that when women are knowledgeable about different modes of treatment they are more inclined to question their rights and demand choices [23]. The satisfaction expressed at all levels of health care could either mean lack of knowledge among women of what care they could expect from antenatal clinics, or lack of difference in quality of care at the various levels of the district referral system.

The results of this study support the findings of others [15,20] that there is an urgent need for improvement of both structural and process quality of antenatal care it may become an effective control program. The poor quality of anaemia management reported in this study may reflect a generally deficient quality also in other important aspects of the current antenatal care program. The high rate of utilization of antenatal services offers an opportunity for the detection and control of anaemia using existing health service contacts. While the main causes of anaemia in this population appear to be malaria, hookworm infestations and iron deficiency, a major weakness in the programme is the diagnostic information and motivation for individual mothers towards specific interventions. Education for pregnant women on what to expect from different health care levels would probably enable them to make choices and therefore understand their levels of satisfaction. While education on quality improvement for health personnel might improve quality of risk detection for antenatal mothers, until then, it is unlikely that the current antenatal program would have any impact in reduction of anaemia in pregnancy.

Address reprint requests to David P. Urassa, Department of Community Health, Muhimbili University College of Health Sciences, P.O. Box 65015, Dar es Salaam, Tanzania. E-mail: Durassa@muchs.ac.tz

We thank Professor Gernard Msamanga of the department of Community Health, Muhimbili University College of Health Sciences, for his guidance during data collection and analysis. We are also grateful to the Research Assistants, District Medical Officer, Health Workers, and the women of the Rufiji district who supported and participated in this study. This study was funded by the Swedish Agency for Research and Co-operation with Developing Countries (SAREC).

References

World Health Organisation.
Mother-Baby Package: Implementing Safe Motherhood in Developing Countries
 . Geneva: WHO, 1994: (http://www.who.int/reproductive-health/publications/MSM_94_11/MSM_94_11_abstract.en.html Accessed July 8
2002
.
Villar J, Bergsjo P. Scientific basis for the content of routine antenatal care. I. Philosophy, recent studies, and power to eliminate or alleviate adverse maternal outcomes.
Acta Obstet Gynecol Scand
 
1997
;
76:
1
–14.
Rooney C.
Antenatal Care and Maternal Health: How Effective is it?
  Geneva: WHO, 1992: http://who.int/dsa/cat98/mat8.htm#Antenatal%20Care Accessed July 8
2002
.
McDonagh M. Is antenatal care effective in reducing maternal morbidity and mortality?
Health Policy Plann
 
1996
;
11:
1
–15.
WHO.
Antenatal Care. Report of a Technical Working Group
 . Geneva: WHO, 1996: http://www.who.int/reproductive-health/publications/MSM_96_8/MSM_96_8_abstract.en.html Accessed July 8
2002
.
Bureau of Statistics.
Tanzania Demographic Health Survey 1996: Dar Es Salaam
 . Calvaton, MD: Bureau of Statistics,
1996
.
Massawe S, Urassa E, Nystrom L, Lindmark G. Effectiveness of primary level antenatal care in decreasing anemia at term in Tanzania.
Acta Obstet Gynecol Scand
 
1999
;
78:
573
–579.
WHO. New estimates for maternal mortality.
Weekly Epidemiol Record
 
1996
;
71:
97
–104.
Urassa EJN, Massawe S, Lindmark G, Nystrom L. Maternal mortality in Tanzania—medical causes are interrelated with socio-economic and cultural factors.
S Afr Med J
 
1996
;
86:
436
–444.
Kielmann AA, Janovsky K, Annett H.
Assessing District Health Needs, Services and Systems: Protocols for Rapid Data Collection and Analysis
 . London: Macmillan and AMREF,
1995
.
UNICEF, WHO, UNFPA.
Guidelines for Monitoring the Availability and Use of Obstetric Services
 , 2nd edn. New York: UNICEF,
1997
.
Altman GD. Some common problems in medical statistics. In:
Practical Statistics for Medical Research
 . London: Chapman and Hall,
1991
, pp.
396
–409.
Massawe S, Urassa E, Lindmark G, Moller B, Nystrom L. Anemia in pregnancy: a major health problem with implications for maternal health care.
Afr J Health Sci
 
1996
;
3:
126
–132.
Bergsjo P, Seha A, Ole-Kingori N. Hemoglobin concentration in pregnant women: experience from Moshi, Tanzania.
Acta Obstet Gynecol Scand
 
1996
;
75:
241
–244.
WHO.
The Prevalence of Anemia in Women
 , 2nd edn. Geneva: WHO,
1992
.
Meda N, Cousens S, Kanki B. Anemia among women of reproductive age in Burkina Faso.
World Health Forum
 
1996
;
17:
369
–372.
Fleming AF. Tropical obstetric and gynaecology. Anemia in pregnancy in tropical Africa.
Trans R Soc Trop Med Hyg
 
1989
;
83:
441
–448.
Premji Z, Makwaya C, Minjas JN. Current clinical efficacy of chloroquine for the treatment of Plasmodium falciparum infections in urban Dar Es Salaam, United Republic of Tanzania.
Bull World Health Organ
 
1999
;
77:
740
–744.
Ministry of Health Tanzania.
National Guidelines on Malaria Diagnosis and Treatment
 . Dar Es Salaam: Ministry of Health Tanzania,
2000
.
Van den Broek NR, Ntonya C, Mhango E, White SA. Diagnosing anemia in pregnancy in rural clinics: assessing the potential of the Hemoglobin Colour Scale.
Bull World Health Organ
 
1999
;
77:
15
–21.
Ekstrom EC, Kavishe FP, Habicht JP, Frongillo EAJ, Rasmussen KM, Hemed L. Adherence to iron supplementation during pregnancy in Tanzania: determinants and hematologic consequences.
Am J Clin Nutr
 
1996
;
64:
368
–374.
Le Roux E, Pattinson RC, Tsaku W, Makin JD. Does successful completion of the Perinatal Education Programme result in improved obstetric care?
S Afr Med J
 
1994
;
88:
180
–182.
Kabakian-Khasholian T, Campbell O, Shediac-Rizkallah M, Ghorayeb F. Women’s experiences of maternity care: satisfaction or passivity?
Soc Sci Med
 
2000
;
51:
103
–113.