Management of severe acute malnutrition by cow milk in resource constraints settings: experience of the Nutritional Centre of the University Clinics of Graben

The study was conducted in the Paediatric Service of the University Hospital of Graben in the UNTI Giorgio Cerruto, with a capacity of 30 beds. The Centre admits an average of 22 cases of severe malnutrition monthly.

We conducted a retrospective comparative study on the records of patients who were followed up for SAM. Patients with complete records were included in the analysis. Those with incomplete records were excluded. 136 patients met the inclusion criteria and constitute the study sample. Of these, 57 patients had received therapeutic milk F75, F100 and RUTF (LT / RUTF group) for a three months period prior to the stockout (January–March 2015) and 79 were administered cow milk / vegetal oil-sugar-maize-soybean (CM/MASO group) for three months (April–June 2015) when the UNTI was out of stock in conventional therapeutic supplies.

Data were collected from the individual patient records and entered in an electronic database in SPSS 20. For each patient, the following variables were considered: age, sex, nutritional diet, weight, size, presence of oedema, signs of intolerance, diagnosis, comorbidity and clinical outcome.

We compared weight changes, oedema resolution at Day (D) 1 (baseline), D3, D5, D7, D14 and D21, the tolerance of the respective regimens and the clinical outcome in both groups.

Weight was measured in kilograms; oedema resolution in number of crosses (3 crosses for generalized oedema; two crosses for bilateral oedema of the feet, ankles and legs, hands and forearms; one cross for the bilateral feet oedema [6] and tolerance by the occurrence of diarrhoea, vomiting or constipation in the week following the initiation of nutritional diet [10]. Clinical outcome was assessed as “improved”, “not improved” or “died”, based on attending physician’s judgement which was recorded on patients’ files.

Data were checked for consistency and completeness by matching the source records, the data collection form and the electronic database by a double independent verification process and were analysed using SPSS 20 software.

Our study protocol was approved by the Ethics Committee of North-Kivu. As we only searched data from the records of previously treated patients, risk was deemed minimal for them. All personal identification information was coded in the database. Source documents were accessible only by assigned and authorized staff as the research team routinely abides to the ethical duty of medical confidentiality.

We analysed the records of a total of 136 patients including 65 females (47.8%) and 71 males (52.2%). In these, 96 (70.6%) had kwashiorkor; 22 (16.2%) marasmus and 18 (13.2%) mixed severe acute malnutrition. The median age was 24 months for the CM / MASO group (minimum age 8 months; maximum age 204 months); 36 months for the TM / RUTF group (minimum age 3 months, maximum age 144 months). At baseline, median weight was 8.4 kg in the cohort CM / MASO (minimum 4.1 kg; Maximum 17 kg); 9.4 kg cohort in TM / RUTF (minimum 1.7 kg; maximum 18 kg) (Table 1).

This section gives the comparison of means of weight in each type of malnutrition and nutritional management. The evolution is given at D1, D3, D5, D7, D14 and D21. The results for patients with Kwashiorkor are showed in Table 2 with its graph in Fig. 1; the ones for patients with marasmus in the Table 3 and Fig. 2 and for the patients with mixed malnutrition in the Table 4 and Fig. 3.

Table 2 shows that patients with Kwashiorkor in CM / MASO group gained an average of 0.8 Kg from D1 (9.73 Kg) to D21 (10.53 Kg). In the TM/ RUTF group, the mean weight rose from 10.07 Kg at D1 to 10.51 kg at D21 (mean weight gain 0.43 kg). After the Anova test, the observed differences were not statistically significant. The calculated F test were less than the tabular F test with P-value more than 0.05.

Table 3 shows that children with marasmus in the CM / MASO group gained a mean weight of 0.97 kg from D1 (6.14 kg) to D21 (7.11 kg). In the TM / RUTF group, the average weight did not change from D1 to D21.

Table 4 shows that the children with mixed malnutrition in the CM / MASO group had a mean of 7 kg at D1 and 7.55 kg at D21 which represented a mean weight gain of 0.55 kg. In the TM / RUTF group, the mean weight was 9,75Kg at D1, 10,33 kg at D21 which represented a weight gain of 0,58 kg. After the Anova test, these differences were not significant. The P values were all more than 0.05.

Kwashiorkor or mixed malnutrition are recognized for bilateral oedema which may be severe (+++), moderate (++) or mild (+).

In both groups, oedema resolved the same way (Table 5, Fig. 4). On day 7, half of initially oedematous patients had recovered. On D14, four patients still remained oedematous in the CM / MASO cohort. Their comorbidities were: gastroenteritis (two), pneumonia and intestinal parasitosis (one); urinary tract infection and intestinal parasitosis (one). In the TM / RUTF cohort, two patients were still oedematous. Their comorbidities were: intestinal parasitosis (one); tuberculosis, chronic enteritis and trophic ulcers on a psychomotor retardation background (one). On D21, no patient was oedematous. We conducted a Student t test to check the homogeneity of the two groups. The calculated T is 0.950 with 1.96 as T theoretical. No significant difference between the two treatment groups was found.

Tolerance of regimen was defined by the occurrence or not in the week following the initiation of the nutritional management of one or more of the following symptoms: diarrhoea, vomiting and constipation (Table 6).

We recorded two fatal cases (one in each group); two patients did not improve in the CM/MASO cohort. They presented the following comorbidities: gastroparesis (one), severe cerebral palsy (one). There was no significant difference in the two groups regarding clinical outcome.

Nutritional rehabilitation is based on weight control until a weight-for-height z-score > − 1 is reached or the MUAC is > 125 mm [1]. When we compared the averages of weight change in both cohorts regarding each type of malnutrition, mean weight gain was similar. Kwashiorkor patients first dropped down due to oedema resolution before increasing their weight from D7 while the patients with marasmus and mixed malnutrition had stable weight the first week before increasing the following weeks. We are not able to confirm that children recovered normal weight since data of weight-for-height z-scores and MUAC were missing for D21 and the period considered for analysis was too short as WHO recommends that children with severe acute malnutrition should only be discharged from treatment when their weight-for-height is ≥ − 2 Z-score and the MUAC is ≥125 cm and they had no oedema for at least 2 weeks [6]. Therefore, the follow up period for SAM management can go up to six weeks for full recovery. Nguefack and collaborators presented similar results in their study conducted at the University Hospital of Yaounde on the hospital management of severe acute malnutrition in children with local preparations made from whole milk (Nido®), cooking oil, sugar, cereals and multivitamins [12]. In Basra (Iraq), Sawsan I. Habeeb also found similar results in a retrospective study that evaluated the therapeutic effectiveness of F75 and F100 prepared locally (dried whole milk, cereals flour, sugar, vegetable oil and minerals) in comparison with industrially processed milk [13].

Nutritional oedema is a common symptom in kwashiorkor and mixed malnutrition. It predicts for severity [5, 6, 14‐16]. Their resolution is therefore a good surrogate for a positive response to nutritional management [17].

In both groups in our study, oedema resolved at a similar rate (Fig. 4). At the end of the first week of treatment, half of oedematous patients had resolved. At the end of the second week, six patients (four in the cohort CM / MASO and two in the TM / RUTF cohort) were still oedematous and at the end of the third week, no patient presented oedema. The six who still retained oedema at D14 had the following comorbidities: pneumonia, gastroenteritis, digestive parasitosis, urinary tract infection, cerebral palsy, sepsis and tuberculosis. Dominique Roberfroid and his collaborators in a systematic review of oedematous malnutrition management had also found that concomitant infections are among the risk factors of mortality and treatment response delay [17].

While diarrhoea is a common symptom and a criterion of severity in SAM [18, 19], it is also a sign of digestive intolerance due to malabsorption alongside vomiting and constipation [11]. In both cohorts, tolerability was similar (p > 0.05) characterized by diarrhoea in a quarter of patients and vomiting in less than a fifth. Constipation was rare. Razafindrakoto and his colleagues in their study that compared goat with cow milk in the management of SAM also found that patients did not encounter digestive intolerance with cow milk [10].

We recorded two deaths (1.5% total mortality rate, 1.3% in the CM / MASO group; 1.8% in the TM / RUTF group). The overall improvement rate was 97% (CM / MASO 96.3%; TM / RUTF 98.2%). Lower recovery rates (33.6%) were found in Tamale Teaching Hospital (Ghana) in a retrospective chart review study [20]. And a bigger mortality rate (3.7%) was recorded in Niger when evaluating a nutritional rehabilitation program, in patients with severe malnutrition treated according to the WHO standard procedures [21] while it ranged from 3.4 to 35% in a systematic review and meta-analysis of the management of severe acute malnutrition in low and middle income settings [22]. The reported high improvement and low mortality rates in our study may be due to the short study period.

Data on patients’ MUACs and weight for height z-scores were missing for D21. So we were unable to assess patients’ clinical outcome objectively. To address this limitation, we relied on the physician’s final judgement at patient’s discharge which was recorded on patients’ files. We were also unable to present a thorough discussion of the results because not many research works have been published on the comparison between WHO recommended therapeutic milks and locally processed therapeutic foods to treat SAM.