Abstract

Aim: To assess the clinical outcomes of a combined approach to the treatment of severe acute malnutrition in an area of high HIV prevalence using: (i) an initial inpatient phase, based on WHO guidelines and (ii) an outpatient recovery phase using ready-to-use therapeutic food.

Methods: An operational prospective cohort study implemented in a referral hospital in Southern Malawi between May 2003 and 2004. Patient outcomes were compared with international standards and with audits carried out during the year preceding the study.

Results: Inpatient mortality was 18% compared to 29% the previous year. Programme recovery rate was 58.1% compared to 45% the previous year. The overall programme mortality rate was 25.7%. Of the total known HIV seropositive children, 49.5% died.

Conclusions: Inpatient mortality and cure rates improved compared to pre-study data but the overall mortality rate did not meet international standards. Additional interventions will be needed if these standards are to be achieved.

Introduction

Child malnutrition is a significant global problem, accounting for up to 56% of 10.6 million child deaths annually in developing countries [1]. Severe acute malnutrition (SAM) is commonly associated with case fatality rates of up to 60% in treatment units [2, 3]. Malawi, like many countries in sub-Saharan Africa, faces the added burden of an HIV epidemic [4].

In a hospital-based HIV prevalence study in Southern Malawi, 40% of the malnourished children tested HIV positive and HIV infection contributed to over 40% of all paediatric deaths [5]. HIV sero-prevalence among malnourished children was reported as 34.4% and overall inpatient Nutrition Rehabilitation Unit (NRU) mortality was 28% [6]. The inpatient case fatality rate for children with kwashiorkor was 30.5% [7]. Difficulties in accessing food and medication supplies and in staff training and motivation were identified.

In the same hospital NRU research using peanut-based Ready-to-Use Therapeutic Food (RUTF) had demonstrated good weight gains and cure rates of 86 and 75%, respectively for HIV-negative and positive children [8, 9]. However, these encouraging results were recorded in children recruited after stabilization and did not address the ongoing high inpatient mortality. These studies were able to remunerate travel costs to enable outpatient clinic attendance through to recovery.

Following the declaration of a National Food Emergency in 2002, the NRU started to receive international NGO support and to record outcome data in line with internationally accepted nutrition standards. Staffs were retrained to WHO SAM guidelines [10], locally produced feeds were replaced by pre-packed F75 and F100 formulae milks (Nutriset, France) and supply lines of drugs, blankets and bed nets were established. An audit from December 2002 to March 2003 showed that the inpatient case fatality rate remained unchanged at 29% (102/350), with HIV/AIDS infection contributing to 32% of deaths. Average inpatient stays were 14 days with cure rates of 45% (158/350) (Geenan unpublished results, 2003).

The Malawi Ministry of Health and Population shared concern about the continuing poor outcomes and agreed that a new treatment approach be evaluated.

Aim

To assess the clinical outcomes, primarily recovery and case fatality rates, of a novel combined approach to the treatment of SAM using:

  • an initial inpatient stabilization phase, based on WHO guidelines,

  • a subsequent outpatient recovery phase with a locally produced RUTF.

Methodology

This was an operational prospective cohort study. Ethical approval was obtained from the Ministry of Health and Population, Malawi and from Great Ormond Street Hospital for Sick Children NHS Trust/Institute of Child Health Research Ethics Committee, UK.

Study site and staff

Moyo House NRU is situated within Queen Elizabeth Central Hospital, Blantyre. Blantyre District is 2012 km2 and has a population of 809 397 of whom 502 053 live in urban and semi-urban township areas, the remainder in rural areas [11]. The estimated prevalence of HIV among people aged between 15 and 49 years in Blantyre District is 22.3%, the highest in the country [4]. Admissions to the NRU are on average 1200 per year. At the time of the study HIV treatment programmes were not well established.

NRU staff included six full time nurses, two home craft workers and two patient attendants, supported by an experienced paediatrician and 2–3 trainee paediatricians.

Staff received training from WHO trainers on the inpatient management of SAM, with additional training in RUTF outpatient protocols.

Subjects

This study analysed data from 1237 children admitted for treatment to Moyo House, between 23 May 2003 and 22 May 2004.

Criteria for inclusion in analysis were:

  • SAM according to WHO classification [weight for height (WFH) ≤70% or –3 z-scores of the NCHS reference and/or bilateral pitting oedema] [12]

  • age more than 6 months

  • absence of severe disability

NRU inpatient protocols

On admission children received standard inpatient phase 1 treatment according to WHO-based Malawi National Guidelines for the treatment of SAM. These included an initial phase of Formula 75 milk at 100 kcal kg–1 per day followed by a transition phase and then phase 2 feeds with Formula 100 milk at 200 kcal kg–1 per day [10]. Children were weighed daily and received daily clinical review.

Medication was also given according to national SAM treatment protocols. Additional laboratory investigations were undertaken on the basis of clinical need.

HIV counselling and testing was intermittently available for children and carers, and was carried out mainly according to clinical indication. Routine testing was not available.

Outpatient protocols

Children were discharged from the NRU once clinically stable and tolerating phase 2 feeds. They received outstanding medication, a ration of RUTF (170 kcal kg–1 per day) and a family ration of 2 kg of a blended maize-soya flour with oil (700 kcal per day).

At fortnightly follow-up clinics children were nutritionally and medically assessed and received a further 2 week RUTF and family ration. Discharge followed nutritional cure, clinical stability and freedom from treatable infection.

Children who defaulted from the programme were visited at home.

Data collection

Admission and follow-up weights and heights were taken with calibrated standard ‘Salter’ spring scales accurate to 100 g and locally constructed height boards accurate to 0.5 cm. All measurements including medical complications and the presence of bilateral pitting oedema were recorded on admission and at follow-up on a standard individual treatment card.

Coding and analysis

Individual record data on anthropometry, length of stay, oedema, HIV status and outcome were transferred into Epiinfo version 6.04 [13] using the Epidata entry programme [14].

Outcome data was coded as ‘recovered’, ‘died’, ‘transferred out’ or ‘defaulted’.

‘Recovery’ or nutritional cure was defined as reaching >85% of the reference median WFH.

‘Defaulters’ were those who failed to attend the clinic on two consecutive occasions. If they were subsequently found to have died within 1 month of their last outpatient clinic attendance they were re-coded as ‘deaths’.

Individual weight gains in marasmic patients were calculated using the admission weight, weight gains in oedematous patients were calculated using the weight at resolution of oedema. Missing data values were coded as missing.

Outcomes were compared with international standards for therapeutic care. Proportions were compared using the chi-squared test with Yates continuity correction.

Carer survey

An independent carer survey approved by the College of Medicine Research and Ethics Committee, collected data by interviews and focus group discussions (FGDs). Respondents were chosen randomly. The interviews were conducted privately, using a structured questionnaire and lasted for 20–30 min. Notes were made by a local research assistant. Each FGD included 5–6 carers, lasted for 45–60 min and was cassette-recorded and translated before analysis (M Ageyman MSc Trop Paeds Dissertation, Liverpool School of Tropical Medicine).

Results

Admission WFH data were recorded for 1044 and follow-up weight change data for 791 of the 1077 admissions analysed in this study. A further 193 children were treated but did not fulfil study admission criteria.

Table 1 presents patient admission profile. Sex distribution was equal (P > 0.05), with 53.7% males and 46.3% females. A total of 843 children (78.3%) suffered from oedematous malnutrition [732 (68.0%) kwashiorkor and 111 (10.3%) marasmic-kwashiorkor], and 228 (21.2%) from marasmus.

Table 1

Characteristics at admission

Variable Total
 
Marasmus
 
Marasmic kwashiorkor
 
Kwashiorkor
 
>6 month <4 kg
 
Age (months) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) 
Overall 1077 24 (17–33) 228 19 (13–25) 111 20 (15–26) 731 25 (19–36) 9 (6–13) 
Recovered 626 24 (18–36) 76 19 (14–26) 42 19 (15–27) 506 25 (19–37) 11 (8–13) 
Died 276 21 (14–28) 108 20 (13–27) 47 18 (15–28) 117 24 (16–32) 8 (6–12) 
Defaulted 174 23 (16–28) 44 18 (12–24) 22 23 (15–25) 108 24 (18–36)  
 
Admission weight for height (z-scores) 
 n Mean (SD) n Mean (SD) n Mean (SD) n Mean(SD) n Mean (SD) 
 
Overall 1044 −2.3 (1.3) 226 −3.7 (0.6) 111 −3.8 (0.6) 705 −1.7 (1.1) −2.9 (0.1) 
Recovered 614 −2.0 (1.2) 76 −3.5 (0.5) 42 −3.7 (0.6) 495 −1.6 (1.0) −2.8 
Died 260 −3.1 (1.2) 107 −3.8 (0.7) 47 −3.9 (0.6) 105 −2.1 (1.0) −3.0 
Defaulted 170 −2.5 (1.3) 43 −3.7 (0.5) 22 −3.7 (0.4) 105 −1.8 (1.1)  
Variable Total
 
Marasmus
 
Marasmic kwashiorkor
 
Kwashiorkor
 
>6 month <4 kg
 
Age (months) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) 
Overall 1077 24 (17–33) 228 19 (13–25) 111 20 (15–26) 731 25 (19–36) 9 (6–13) 
Recovered 626 24 (18–36) 76 19 (14–26) 42 19 (15–27) 506 25 (19–37) 11 (8–13) 
Died 276 21 (14–28) 108 20 (13–27) 47 18 (15–28) 117 24 (16–32) 8 (6–12) 
Defaulted 174 23 (16–28) 44 18 (12–24) 22 23 (15–25) 108 24 (18–36)  
 
Admission weight for height (z-scores) 
 n Mean (SD) n Mean (SD) n Mean (SD) n Mean(SD) n Mean (SD) 
 
Overall 1044 −2.3 (1.3) 226 −3.7 (0.6) 111 −3.8 (0.6) 705 −1.7 (1.1) −2.9 (0.1) 
Recovered 614 −2.0 (1.2) 76 −3.5 (0.5) 42 −3.7 (0.6) 495 −1.6 (1.0) −2.8 
Died 260 −3.1 (1.2) 107 −3.8 (0.7) 47 −3.9 (0.6) 105 −2.1 (1.0) −3.0 
Defaulted 170 −2.5 (1.3) 43 −3.7 (0.5) 22 −3.7 (0.4) 105 −1.8 (1.1)  

Table 2 presents outcome data stratified by HIV status. Overall recovery rate was 58.1% (CI =55.1–61.0); case fatality rate 25.7% (CI = 23.1–28.3) and default rate 16.2% (CI = 14.0–18.4).

Table 2

Outcome by HIV status

 Total
 
HIV +
 
HIV−
 
Unknown
 
Outcome n (%) n (%) n (%) n (%) 
    Overall 1077  186  73  818  
    Recovered 626 58.1 64 34.4 50 68.5 512 62.6 
    Died 277 25.7 92 49.5 11 15.1 174 21.3 
    Defaulted 174 16.2 30 16.1 12 16.4 132 16.1 
 
Length of stay (days) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) 
 
    Overall 1059 38 (17–62) 185 41 (13–85) 73 49 (27–82) 801 37 (17–55) 
    Recovered 626 49 (36–70) 64 69 (50–109) 50 63 (41–84) 512 47 (36–64) 
    Died 260 7 (2–18) 91 15 (7–49) 11 10 (7–23) 158 5 (2–11) 
    Defaulted 173 24 (13–50) 30 36 (24–100) 12 31 (20–83) 131 21 (10–41) 
 Total
 
HIV +
 
HIV−
 
Unknown
 
Outcome n (%) n (%) n (%) n (%) 
    Overall 1077  186  73  818  
    Recovered 626 58.1 64 34.4 50 68.5 512 62.6 
    Died 277 25.7 92 49.5 11 15.1 174 21.3 
    Defaulted 174 16.2 30 16.1 12 16.4 132 16.1 
 
Length of stay (days) n Median (IQ range) n Median (IQ range) n Median (IQ range) n Median (IQ range) 
 
    Overall 1059 38 (17–62) 185 41 (13–85) 73 49 (27–82) 801 37 (17–55) 
    Recovered 626 49 (36–70) 64 69 (50–109) 50 63 (41–84) 512 47 (36–64) 
    Died 260 7 (2–18) 91 15 (7–49) 11 10 (7–23) 158 5 (2–11) 
    Defaulted 173 24 (13–50) 30 36 (24–100) 12 31 (20–83) 131 21 (10–41) 

A total of 259 children (24.0%) were tested for HIV. Of these, 186 children (71.8%), 17.3% of the whole study population were confirmed HIV seropositive. HIV seropositivity was significantly higher (P < 0.0001) in marasmic children (90/228, 39.5%) than in those suffering from oedematous malnutrition (94/842, 11.2%).

HIV had significant impact on mortality. Of the total known HIV-seropositive children, 92/186 (49.5%) died vs. 11/73(15.1%) of known seronegative children (P < 0.001). Of outpatient deaths 46/83 (55.4%) were confirmed HIV positive. Of the overall case fatality rate (25.7%), 194/1077 (18.0%) died whilst inpatients and 83/883 (9.4%) died as outpatients.

A total of 103 (53.1%) of the inpatient deaths occurred within the first 4 days of admission (Fig. 1).

Fig. 1.

Kaplan-Meier time event analysis of inpatient deaths.

Fig. 1.

Kaplan-Meier time event analysis of inpatient deaths.

Of the 174 defaulters 84 (48.3%) were traced and a final outcome recorded (Fig. 2). The recovery rate rose to 63% when defaulted children who had nutritionally recovered by follow-up were recoded as recovered.

Fig. 2.

Defaulter outcomes n = 174.

Fig. 2.

Defaulter outcomes n = 174.

Overall, median (IQ range) length of stay in the programme for those children that recovered was 49 days (36–70). Length of stay was longer for confirmed HIV-seropositive children [69 days (50–109)] than for HIV-seronegative children [63 days (41–84)] but not significantly so (Table 2). The median (IQ range) inpatient stay was 9 days (7–13) and outpatient stay was 36 days (28–59).

Overall, the median (IQ range) weight gain in the programme for children that recovered was 5.2 g kg–1 per day (3.4–7.5).

Figure 3 compares inpatient mortality collected by the hospital HMIS (Health Management and Information System) during corresponding months of 2003–2004 (the year of this study) and 2002–2003, (the year immediately before this study). This independently collected data indicated a downward trend in mortality in the unit. This was also shown by comparison of the December 2002 to March 2003 audit data, which showed an inpatient mortality of 29% (102/350), with the study inpatient mortality of 18% (626/1077) (χ2 = 17.5 P < 0.0001). Comparison of the rate of recovery and the inpatient length of stay seen in this study with that observed during the audit also indicates improvement (58.1% vs. 45% and 9 days vs. 14 days, respectively).

Fig. 3.

MOYO inpatient deaths pre- and post-RUTF (HMIS data).

Fig. 3.

MOYO inpatient deaths pre- and post-RUTF (HMIS data).

FGD identified that RUTF was widely accepted by carers and children including those with advanced HIV disease. The shorter inpatient stay and follow-up clinics were valued by carers. The main reasons for defaulting from follow-up were prohibitive transport costs and long walking distances to the central NRU.

Discussion

Compared to pre-intervention HMIS and audit data, this study improved recovery rates, reduced inpatient stays and reduced inpatient mortality. The findings of FGDs confirmed the acceptability of the programme to carers. However, the overall recovery rate and mortality rate failed to reach international Sphere standards [15].

The high HIV prevalence in this setting is almost certainly an important contributing factor to this persistently high mortality. In 2000 Rogerson, et al. [5] found that HIV infection contributed to over 40% of all paediatric deaths and in 2005 Kerac, et al. found that HIV contributed to over 60% of deaths among children suffering from SAM (Kerac M, Bunn J. ‘Excess mortality risk associated with HIV infection in a large Malawian NRU’ 2nd Malawi Annual HIV Nutrition Meeting, 27 January 2006, unpublished results). In our study, 33.2% of inpatient deaths occurred in patients known to be HIV positive. However, we had only limited access to HIV testing facilities and had to prioritize testing on the basis of clinical need. It is likely that some of the mortality in the group with unknown HIV status was also HIV related. Of the children who died in outpatient care, 55.4% were confirmed HIV seropositive. The programme mortality rate for confirmed HIV-seronegative children was 15.1%, not significantly higher than the Sphere minimum standard of 10% (95% CI 6.9–23.3 P = 0.3). At the time of the study, access to HIV treatment programmes for both adults and children was restricted to those able to afford medication and to the small number of children resident in areas served by NGO funded programmes. Similarly PMTCT programmes were not universally in place. Routine testing and ARV treatment are now increasingly available and linkages between the NRU and treatment teams are being established, this may impact on the mortality rates reported here. Those HIV-positive children that did recover in this study took 20 days longer to do so than HIV-negative children. In this setting, RUTF proved particularly useful in allowing their slower recovery at home.

A total of 53.1% of inpatient deaths occurred within 4 days of admission (Fig. 2), suggesting that children were very sick on admission. The severity of SAM and its prognosis are closely related to the lead-time to presentation [1, 3, 7, 16–18]. As Moyo House is a centralized unit covering a large catchment area, it is likely that distance and high opportunity costs of travelling to and staying at the unit delayed presentation and is one factor that contributed to this mortality. This impression was supported by the qualitative data collected by FGD and interviews with the mothers of malnourished children receiving treatment. This indicated that, although carers appreciated the reduced length of inpatient stay, they wanted to be able to access care closer to their homes and have fewer opportunity costs associated with care. Over 70% (63/86) of traced defaulters were well and of those that specified a reason for default 45% (9/20) cited high opportunity costs as the main reason for not completing treatment.

Although Moyo house is relatively well resourced compared to other hospitals in Malawi [7], staffing levels were barely sufficient to implement basic elements of the WHO ‘10 Steps’ protocols [19]. Especially in the rainy season, nurse to patient ratios were commonly 1 : 50 and optimal individual patient care and/or monitoring was problematic. This finding reflects similar experience elsewhere in Africa where lack of human resources, in particular skilled staff, has been identified as a major obstacle in reducing case fatality rates and is a possible contributor to the high mortality, especially that seen among HIV-negative children, seen here [2, 19, 20].

The outpatient treatment protocols used in this study differ from protocols for the community-based management of acute malnutrition (CMAM) endorsed by a joint UN statement [21]. In CMAM, access to services is decentralized, the majority of cases of SAM are treated solely as outpatients and only those children suffering from additional medical complications are admitted into inpatient care [22]. In contrast, the treatment in this study remained centralized and all cases of SAM were initially admitted as inpatients. This placed high demands on families accessing care and may be a factor that contributed to delayed presentation. Allowing families to access care in their local clinic could reduce their opportunity costs, encourage earlier presentation and thereby reduce the proportion who present with medical complications. A pilot site that provided outpatient care from one rural clinic during this study was welcomed by carers and showed low default and high cure rates. This CMAM approach has succeeded in rural areas of Malawi that have lower HIV prevalence rates [23], and in programmes that recruited children already stabilized [24]. A decentralized CMAM intervention in a high HIV prevalence area would need to ensure access to HIV and other diagnostic testing and clinical treatment facilities using well-defined referral criteria.

In conclusion, the successful management of SAM in high prevalence HIV areas is likely to require a combination of interventions including linkage to HIV support, testing and treatment services, decentralized outpatient treatment with RUTF for all children with uncomplicated SAM and provision of timely and specialized paediatric inpatient care for children with severe concurrent infections. The value of these additional interventions will need further assessment.

Acknowledgements

This work was undertaken as part of the Community Therapeutic Care Programme, a collaboration between Valid International and Concern Worldwide. It was generously supported by Matthew and Sybil Orr, the World Health Organisation; the United States Agency for International Development (USAID) through the Food and Nutrition Technical Assistance (FANTA) project/Academy for Educational Development (AED); Ireland Aid and Brixton Health. We would also like to thank the staff, parents, carers and children at Moyo House. Without their support and participation this study would not have been possible.

References

1
Pelletier
DL
Frongillo
EA
Jr
Schroeder
DG
, et al.  . 
The effects of malnutrition on child mortality in developing countries
Bull World Health Organ
 , 
1995
, vol. 
73
 (pg. 
443
-
8
)
2
Schofield
C
Ashworth
A
Why have mortality rates for severe malnutrition remained so high?
Bull World Health Organ
 , 
1996
, vol. 
74
 (pg. 
223
-
9
)
3
Collins
S
Dent
N
Binns
P
, et al.  . 
Management of severe acute malnutrition in children
Lancet
 , 
2006
, vol. 
368
 (pg. 
1992
-
2000
)
4
National Statistical Office ZM
Demographic and Health Survey. Maryland, USA:
ORC Macro
 , 
2004
5
Rogerson
SR
Gladstone
M
Callaghan
M
, et al.  . 
HIV infection among paediatric in-patients in Blantyre, Malawi
Trans R Soc Trop Med Hyg
 , 
2004
, vol. 
98
 (pg. 
544
-
52
)
6
Kessler
L
Daley
H
Malenga
G
, et al.  . 
The impact of the human immunodeficiency virus type 1 on the management of severe malnutrition in Malawi
Ann Trop Paediatr
 , 
2000
, vol. 
20
 (pg. 
50
-
6
)
7
Brewster
DR
Manary
MJ
Graham
SM
Case management of kwashiorkor: an intervention project at seven nutrition rehabilitation centres in Malawi
Eur J Clin Nutr
 , 
1997
, vol. 
51
 (pg. 
139
-
47
)
8
Manary
MJ
Ndkeha
MJ
Ashorn
P
, et al.  . 
Home based therapy for severe malnutrition with ready-to-use food
Arch Dis Child
 , 
2004
, vol. 
89
 (pg. 
557
-
61
)
9
Ndekha
MJ
Manary
MJ
Ashorn
P
, et al.  . 
Home-based therapy with ready-to-use therapeutic food is of benefit to malnourished, HIV-infected Malawian children
Acta Paediatr
 , 
2005
, vol. 
94
 (pg. 
222
-
5
)
10
WHO. Management of Severe Malnutrition: A Manual for Physicians and Other Senior Health Workers
1999
Geneva
WHO
11
National Statistical Office ZM
Population and Housing Census
 , 
1998
Zomba, Malawi
Ministry of Health and Population
12
National Centre for Health Statistics (NCHS)
NCHS growth curves for children birth to 18 years
1979
Washington: US
Department of Health, Education and Welfare, Government Printing Office
13
EpiInfo version 6: A word processing, database, and statsitics program for public health on IBM-compatible microcomputers
1995
Atlanta, USA
Centres for Disease Control and Prevention
14
EpiData, version 2.1
An Extended Tool for Validated Entry and Documentation of Data
2001
Odense Denmark
The EpiData Association
15
SPHERE Project Team
The Sphere Project: Humanitarian Charter and Minimum Standards in Disaster Response
2004
2 edn
Geneva
The Sphere Project
16
Chen
LC
Chowdhury
A
Huffman
SL
Anthropometric assessment of energy-protein malnutrition and subsequent risk of mortality among preschool children
Am J Clin Nutr
 , 
1980
, vol. 
33
 (pg. 
1836
-
45
)
17
Brewster
D
Improving quality of care for severe malnutrition
Lancet
 , 
2004
, vol. 
363
 (pg. 
2088
-
89
)
18
Briend
A
Dykewicz
C
Graven
K
, et al.  . 
Usefulness of nutritional indices and classifications in predicting death of malnourished children
Br Med J (Clin Res Ed)
 , 
1986
, vol. 
293
  
373–5
19
Karaolis
N
Jackson
D
Ashworth
A
, et al.  . 
WHO guidelines for severe malnutrition: are they feasible in rural African hospitals? Arch Dis Child
2006
20
Ashworth
A
Chopra
M
McCoy
D
, et al.  . 
WHO guidelines for management of severe malnutrition in rural South African hospitals: effect on case fatality and the influence of operational factors
Lancet
 , 
2004
, vol. 
363
 (pg. 
1110
-
5
)
21
WHO, WFP, UNSCN, UNICEF
Community-based Management of Severe Acute Malnutrition: A Joint Statement by the World Health Organization, the World Food Programme, the United Nations System Standing Committee on Nutrition and the United Nations Children's Fund
(1-5-2007, date last accessed) 
22
Collins
S
Sadler
K
Dent
N
, et al.  . 
Key issues in the success of community-based management of severe malnutrition
Food Nutr Bull
 , 
2006
, vol. 
27
 (pg. 
S49
-
82
)
23
Collins
S
Community-based therapeutic care – a new paradigm for selective feeding in nutritional crises
Network paper No 48
 , 
2004
London
Overseas Development Institute. HPN Network Papers
 
Humanitarian Policy Network (ed)
24
Linneman
Z
Matilsky
D
Ndekha
M
, et al.  . 
A large-scale operational study of home-based therapy with ready-to-use therapeutic food in childhood malnutrition in Malawi
Matern Child Nutr
 , 
2007
, vol. 
3
 (pg. 
206
-
15
)

Comments

0 Comments