Abstract

Identification of simple signs and symptoms that predict severe illness needing referral for admission of young infants is critical for reducing mortality in developing countries. Infants <2 months of age presenting to two hospitals in La Paz, Bolivia (n=1082) were evaluated by nurses for signs and symptoms, and independently by physicians for the need for admission. In young neonates, sensitivity of individual clinical signs was >35% for measured temperature ≥37.5°C (65%); all signs had specificity >85%. Odds ratios (ORs) for association of individual clinical signs with need for urgent hospital management were highest (>5) for history of difficulty feeding, not feeding well and fever. Clinical signs or symptoms are useful for primary healthcare workers to identify young infants with serious illness needing admission, and have been incorporated into the Integrated Management of Childhood Illness algorithm for use in Bolivia and elsewhere in Latin America.

Introduction

Infant mortality in Bolivia is 54/1000 and neonatal mortality is 27/1000, among the highest for countries in the America–Caribbean region [1]. Reduction of infant and neonatal mortality and morbidity is a high priority for Bolivia. The main causes of neonatal mortality are infection, birth asphyxia and injury, prematurity and hypothermia [2]. Low birth-weight (LBW) is a common underlying condition in this group [3]. Nearly 70% of neonatal deaths in Bolivia take place in the first week of life [3–5].

Implementation of essential practices, including prenatal care, safe delivery, promotion of breastfeeding, adequate care of LBW babies, and prevention and treatment of asphyxia, infections, hypothermia and hypoglycemia will substantially decrease the number of deaths in young infants [6–13]. As a systematic way of providing essential health care, Bolivia adapted and began to implement Integrated Management of Childhood Illness (IMCI) guidelines in 1997, and has further developed and implemented algorithms for infants <7 days of age since 2002 [14–17]. Little evidence exists, however, for the validity of signs and symptoms that predict severe illness warranting admission to the hospital for infants <60 days of age and particularly those <7 days of age [18–20].

In a previous Young Infant Study, the World Health Organization (WHO) identified signs and symptoms that were predictive of severe illness in young infants [21, 22]. The IMCI algorithm that utilized these signs for classification of illness as the basis for management protocols, however, did not include many young neonates in the first week of life. Moreover, besides sepsis, a variety of conditions that may require referral for increased surveillance and/or management were not sufficiently represented in studies to define IMCI algorithms, among them severe jaundice, LBW or poor feeding due to prematurity. Finally, the first Young Infant Study did not include any country in the America–Caribbean region [21, 23]. Thus, a multicenter study was undertaken to define IMCI clinical algorithms for the identification of sick neonates in the first week of life, and to validate existing algorithms for young infants aged 7–59 days [24].

The objective of this study was to identify simple clinical signs that will help primary health care workers globally, including Latin America, to detect newborns with severe illnesses that require admission to the hospital. Development of an IMCI algorithm for use in the first week of life was specifically sought. This article details the findings from two hospitals in La Paz, Bolivia, and discusses implications for use of IMCI for young infants in Latin America.

Methods

Study site

This study was part of a collaborative, multicenter study involving seven sites: Dhaka (Bangladesh), Kumasi (Ghana), Chandigarh and Delhi (India), Karachi (Pakistan), Durban (South Africa) and La Paz (Bolivia) [24]. The study was undertaken in Bolivia in two major pediatric hospitals: Hospital del Niño (HN) and Hospital Materno Infantil (HMI), both located in the city of La Paz and both teaching hospitals affiliated with the School of Medicine of the Universidad Mayor de San Andrés. Although they are referral facilities, the outpatient and emergency departments of the study hospitals serve as primary care centers for a large portion of the population of both La Paz and El Alto cities (∼2 million inhabitants). The area is located at an altitude of 3300–4100 m above sea level. A National Health Security system [‘Seguro Universal Materno Infantil’ (SUMI)] provides funds through the mayor’s office to ensure health care for all children <5 years of age at no cost to the families. HMI is the only pediatric hospital serving the catchment population that belongs to the National Health Security system [Caja Nacional de Salud (CNS)], and is representative of primary care facilities in the country. However, clients who belong to CNS receive a salary, and have relative financial security; thus, they are not representative of the poorest segments of the population. The population that receives care at HN, however, who do not have a fixed income, generally have more severe health and nutritional problems, and are representative of poorer patients in the primary care system. Together, the patient populations of the two hospitals are broadly representative of the Bolivian healthcare system.

The study lasted for 13 months, from October 2003 through October 2004. It was preceded by a pilot phase (September 2003) in which procedures were tested and problems corrected. The protocol was reviewed and approved by the Ethical Review Committee of the Bolivian Medical College and The Department of Teaching and Research of both study hospitals. The protocol was designated as exempt from Institutional Review Board review at Johns Hopkins University. A Steering Committee was comprised of investigators from WHO, Johns Hopkins University and Boston University.

Subjects

At both hospitals, all parents of children <60 days of age who consulted for an illness between 9 am and 3 pm, and did not have any exclusion criteria were asked for informed consent to participate in the study. The patient was excluded if s/he: (i) resided outside the defined study area (to ensure follow-up); (ii) had been enrolled in the study for an earlier episode of illness, or was being seen for a repeat episode of the same illness; (iii) had been hospitalized in the previous 2 weeks (except for delivery); (iv) had an obvious lethal malformation; (v) required immediate cardiopulmonary resuscitation; (vi) was referred from another health facility; or (vii) did not provide informed consent.

Clinical assessment and laboratory procedures

Once informed consent was given, the child was assessed by an auxiliary nurse, ‘person A’, who had at least 1 year of pre-service health training, and was additionally trained on recognition of the clinical signs in IMCI required by the study. Person A interviewed the mother regarding demographic characteristics of the family, specific symptoms of the illness, and examined the child for vital signs and the presence of clinical signs of illness pre-specified on a case record form. The child was then assessed independently by an experienced pediatrician, ‘person B’, who was blinded to the findings of ‘person A’. Person B determined the primary and secondary diagnoses, and whether or not the child needed urgent referral level medical care at the hospital (i.e. admission). Thus, the judgment of the expert pediatrician became the gold-standard outcome for analysis of the sensitivity and specificity of the IMCI signs used by the nurse to identify sick young infants.

All those sent home had a follow-up visit in the hospital within 3 days, or a nurse was sent to the home if the patient did not return, to assess the condition of the child, specifically whether the child was well, improved, sick and needing hospitalization or had died. Children admitted were submitted to all the procedures requested by hospital routine and sound pediatric care practices [25].

Pulse oximetry was performed in all infants (Nellcor, Pleasanton, CA, USA). All those who needed admission underwent a blood culture and serum glucose determination. All jaundiced children had a serum bilirubin determination (Diazo method, Synchron CX system, Beckman Instruments, Inc, USA). Chest radiograph was performed if the admitted child had signs of respiratory illness, and lumbar puncture was done on seriously ill young infants with possible meningitis.

The clinical course and laboratory results of the patient in the hospital were recorded, including the outcome, whether cured, improved, left the hospital against medical advice, or died. Final clinical diagnoses were also recorded.

All study forms were reviewed by the principal investigators, and completeness was ensured. The forms of a random sample of 20% of admitted patients and 10% of those sent home were reviewed on a regular basis by a Review Committee comprised of the two study principal investigators and three other pediatricians from the Bolivian Pediatric Society. In all cases in which the Review Committee had a question regarding the final judgment of study person B on need for hospital admission, they met together to review the case and reach consensus on whether they agreed with the decision of study person B.

Several site visits were made by a member of the multi-country Steering Committee (GLD) to ensure that study procedures were followed according to the protocol.

Data analysis

Data were entered on case record forms specifically designed for the study, and transferred to EpiInfo (version 6b) by two separate data processors to eliminate transcription errors. Data files were sent to the international data coordinating center in Melbourne, Australia, where further consistency checks were performed and data were further cleaned in consultation with the study principal investigators. Data analysis was conducted at the data center using Stata version 8 software (Stata Corporation, College Station, Texas, USA).

Sensitivity and specificity of individual symptoms and clinical signs were determined, based on the gold-standard determination of need for urgent referral-level hospital care. Univariate logistic regression analysis was used to estimate ORs [with 95% confidence intervals (CIs)] for the association of individual symptoms and signs with need for hospital admission. This analysis did not encompass signs of jaundice, which will be the subject of a separate report. Thus, patients admitted for jaundice were included in the non-admitted group for purposes of the sensitivity, specificity and OR analyses reported here.

Results

Presenting complaints and characteristics

A total of 1748 children were triaged, 650 at HMI and 1098 at HN; 1082 infants entered the study, 476 aged 0–6 days and 606 aged 7–59 days. Of those who were excluded, 80% were excluded because they had been referred from another health facility.

Presenting complaints are presented by age group in Table 1; data on presenting complaints were missing for 26.0% (281/1082) of enrolled young infants. Young infants 7–59 days of age were sub-divided into two groups: the late neonatal period, days 7–27; and the second month of life, days 28–59. The single most common complaint in the early and late neonatal periods was jaundice. Umbilical and eye problems (conjunctivitis) were also common in neonates. Respiratory (nasal obstruction, cough) and gastrointestinal complaints (cramps/colic, diarrhea) were more common in the second month of life.

Table 1

Frequency (n, %a) of primary presenting complaints in study infants, by age category

 Age group
 
 0–6 days 7–27 days 28–59 days 
Irritable 16 (4.7) 15 (4.4) 2 (1.6) 
Sleepiness 7 (2.1) 4 (1.2) 1 (0.8) 
Fever 11 (3.3) 7 (2.1) 3 (2.4) 
Cough 3 (0.9) 12 (3.5) 23 (18.4) 
Nasal block/obstruction 9 (2.7) 22 (6.5) 30 (24.0) 
Noisy breathing 1 (0.3) 2 (0.6) 1 (0.8) 
Not feeding well 1 (0.3) 5 (1.5)  
Diarrhea 21 (6.2) 35 (10.3) 12 (9.6) 
Constipation 16 (4.7) 27 (8.0) 8 (6.4) 
Cramps (colic)  12 (3.5) 13 (10.4) 
Mouth problem 2 (0.6) 2 (0.6) 2 (1.6) 
Umbilical problem/discharge 42 (12.5) 41 (12.1) 7 (5.6) 
Eye problem 32 (9.5) 37 (10.9) 10 (8.0) 
Jaundice 167 (49.6) 112 (33.0) 8 (6.4) 
Withdrawal vaginal bleeding 5 (1.5) 1 (0.3)  
Otherb 4 (1.2) 5 (1.5) 5 (4.0) 
Total 337 339 125 
 Age group
 
 0–6 days 7–27 days 28–59 days 
Irritable 16 (4.7) 15 (4.4) 2 (1.6) 
Sleepiness 7 (2.1) 4 (1.2) 1 (0.8) 
Fever 11 (3.3) 7 (2.1) 3 (2.4) 
Cough 3 (0.9) 12 (3.5) 23 (18.4) 
Nasal block/obstruction 9 (2.7) 22 (6.5) 30 (24.0) 
Noisy breathing 1 (0.3) 2 (0.6) 1 (0.8) 
Not feeding well 1 (0.3) 5 (1.5)  
Diarrhea 21 (6.2) 35 (10.3) 12 (9.6) 
Constipation 16 (4.7) 27 (8.0) 8 (6.4) 
Cramps (colic)  12 (3.5) 13 (10.4) 
Mouth problem 2 (0.6) 2 (0.6) 2 (1.6) 
Umbilical problem/discharge 42 (12.5) 41 (12.1) 7 (5.6) 
Eye problem 32 (9.5) 37 (10.9) 10 (8.0) 
Jaundice 167 (49.6) 112 (33.0) 8 (6.4) 
Withdrawal vaginal bleeding 5 (1.5) 1 (0.3)  
Otherb 4 (1.2) 5 (1.5) 5 (4.0) 
Total 337 339 125 

aCalculated as a percent of the total infants reporting a complaint.

bOther: one case of ear problem; two cases of abdominal distension (group 3); two cases of edema (group 1); three cases of vomiting (group 2); one case of respiratory distress in group 2; and two cases in group 1; one case of skin problem in group 2; and two cases in group 3.

General characteristics of the study group are presented in Table 2. The majority of mothers (≥90%) had good antenatal care; most of them (93%) received one or more tetanus toxoid immunizations. Maternal urinary tract infection (42%) and anemia (27%) were relatively common in the antenatal period. Approximately half were first pregnancies. Almost all deliveries (97%) were attended by a skilled provider; the majority (95%) were vaginal. Prolonged labor (13%) and prolonged rupture of membranes (4%) were uncommon. The majority of neonates were exclusively breastfed (94%), but in the second month it decreased to 67%. Neonatal fever, tachypnea, tachycardia and hypoxemia were rare. Average anthropometric z-scores for weight, length and head circumference generally ranged between −0.5 and −1.0. LBW was uncommon (8%).

Table 2

Baseline characteristics of study infants by age groups

 Age group
 
 0–6 days 7–27 days 28–59 days 
Number of infants 476 446 160 
Female (%) 52 48 44 
Exclusive breastfeeding (%) 92 90 67 
Vital signs 
Tachycardia (pulse > 160) (N
Hypoxemia (SaO2 < 85) (N
Anthropometric measurements 
Weight, mean z-score (SD) −1.0 (0.8) −0.9 (0.8) −0.7 (1.1) 
Length, mean z-score (SD) −0.6 (1.0) −0.7 (1.0) −1.3 (1.3) 
Head circumference, mean z-score (SD) −0.6 (0.7) −0.5 (0.7) −0.4 (1.1) 
Obstetric history 
Antenatal care (≥3 visits) (%) 90 93 92 
Parity > 0 (%) 42 49 50 
Maternal tetanus toxoid immunization (≥2 during pregnancy) (%) 88 87 81 
Diabetes during pregnancy (%) 
Anemia during pregnancy (%) 21 26 33 
UTI during pregnancy (%) 42 40 40 
Prolonged labor (%) 13 11 16 
Prolonged rupture of membranes (%) 
Home delivery (%) 
Vaginal delivery (%) 87 71 93 
Delivered by skilled attendant (%) 98 97 94 
Birth weight, kg: mean (SD) 3.2 (0.4) 3.2 (0.4) 3.0 (0.5) 
LBW (<2500 g) (%) 14 
Gestational age (<37 weeks) (%) 
Fever at time of delivery (%) 11 
 Age group
 
 0–6 days 7–27 days 28–59 days 
Number of infants 476 446 160 
Female (%) 52 48 44 
Exclusive breastfeeding (%) 92 90 67 
Vital signs 
Tachycardia (pulse > 160) (N
Hypoxemia (SaO2 < 85) (N
Anthropometric measurements 
Weight, mean z-score (SD) −1.0 (0.8) −0.9 (0.8) −0.7 (1.1) 
Length, mean z-score (SD) −0.6 (1.0) −0.7 (1.0) −1.3 (1.3) 
Head circumference, mean z-score (SD) −0.6 (0.7) −0.5 (0.7) −0.4 (1.1) 
Obstetric history 
Antenatal care (≥3 visits) (%) 90 93 92 
Parity > 0 (%) 42 49 50 
Maternal tetanus toxoid immunization (≥2 during pregnancy) (%) 88 87 81 
Diabetes during pregnancy (%) 
Anemia during pregnancy (%) 21 26 33 
UTI during pregnancy (%) 42 40 40 
Prolonged labor (%) 13 11 16 
Prolonged rupture of membranes (%) 
Home delivery (%) 
Vaginal delivery (%) 87 71 93 
Delivered by skilled attendant (%) 98 97 94 
Birth weight, kg: mean (SD) 3.2 (0.4) 3.2 (0.4) 3.0 (0.5) 
LBW (<2500 g) (%) 14 
Gestational age (<37 weeks) (%) 
Fever at time of delivery (%) 11 

Diagnoses

The frequency of primary diagnoses for admitted and non-admitted patients is presented in Table 3. The majority of patients did not require admission (91%). The proportion of hospitalized patients was greatest for young neonates in the first week of life (14%); only 2.5% of young infants in the second month of life required hospitalization. Overall, the most common primary diagnosis in the first month of life was hyperbilirubinemia/physiological jaundice in both admitted and non-admitted patients; this accounted for >60% of all admissions in neonates. The second most common diagnosis across all age groups was ‘well baby’ (17%). The second most common cause of admission in neonates was suspected sepsis. In the second month of life, sepsis was the main cause of admission but was a rare event (2 in 160). Upper respiratory infection and colic emerged as important problems in the late neonatal period and the second month of life.

Table 3

Frequency of primary diagnosis (pediatrician assessment) by age category and according to whether acute hospital management was recommendeda

 Age group 
 Required admission
 
Did not require admission
 
 0–6 days 7–27 days 28–59 days 0–6 days 7–27 days 28–59 days 
Prematurity/LBW (5) (2) (0) (1) (0) (0) 
Birth injury (0) (0) (0) (2) (1) (1) 
Congenital malformation (2) (2) (0) 10 (2) 11 (3) (6) 
Hyperbilirubinemia/physiological jaundice 42 (64) 27 (61) (0) 194 (47) 165 (41) (4) 
Sepsis 14 (21) (16) (50) (0) (0) (0) 
Pneumonia, ALRI (0) (2) (25) (0) (0) (0) 
URTI, mild ARI (0) (0) (0) (0) 29 (7) 53 (34) 
Skin infection (2) (5) (0) (0) (1) (1) 
Skin condition (non-infectious) (0) (0) (0) (2) (2) (4) 
Eye infection (0) (0) (0) 15 (4) 26 (6) 12 (8) 
Oral thrush (0) (0) (0) (0) (1) (1) 
Umbilical infection, umbilical problem (0) (0) (0) (0) (0) (2) 
Diarrhea, dysentery, persistent diarrhea, dehydration (0) (0) (25) (0) (1) (3) 
Colic (0) (0) (0) (0) 29 (7) 23 (15) 
Feeding problem (0) (9) (0) (1) 13 (3) (2) 
Other (6) (0) (0) (0) (1) 5) 
Well baby (2) (2) (0) 163 (40) 100 (25) 23 (15) 
Total 66 (100) 44 (100) (100) 410 (100) 402 (100) 156 (100) 
 Age group 
 Required admission
 
Did not require admission
 
 0–6 days 7–27 days 28–59 days 0–6 days 7–27 days 28–59 days 
Prematurity/LBW (5) (2) (0) (1) (0) (0) 
Birth injury (0) (0) (0) (2) (1) (1) 
Congenital malformation (2) (2) (0) 10 (2) 11 (3) (6) 
Hyperbilirubinemia/physiological jaundice 42 (64) 27 (61) (0) 194 (47) 165 (41) (4) 
Sepsis 14 (21) (16) (50) (0) (0) (0) 
Pneumonia, ALRI (0) (2) (25) (0) (0) (0) 
URTI, mild ARI (0) (0) (0) (0) 29 (7) 53 (34) 
Skin infection (2) (5) (0) (0) (1) (1) 
Skin condition (non-infectious) (0) (0) (0) (2) (2) (4) 
Eye infection (0) (0) (0) 15 (4) 26 (6) 12 (8) 
Oral thrush (0) (0) (0) (0) (1) (1) 
Umbilical infection, umbilical problem (0) (0) (0) (0) (0) (2) 
Diarrhea, dysentery, persistent diarrhea, dehydration (0) (0) (25) (0) (1) (3) 
Colic (0) (0) (0) (0) 29 (7) 23 (15) 
Feeding problem (0) (9) (0) (1) 13 (3) (2) 
Other (6) (0) (0) (0) (1) 5) 
Well baby (2) (2) (0) 163 (40) 100 (25) 23 (15) 
Total 66 (100) 44 (100) (100) 410 (100) 402 (100) 156 (100) 

ALRI, acute lower respiratory infection; URTI, upper respiratory tract infection.

aRaw frequencies and proportions by column (as percentage, within each category and age group).

The most common secondary diagnoses in neonates were feeding problem, congenital malformation and diarrhea (Tables 3 and 4).

Table 4

Frequency of all diagnoses (pediatrician assessment) by age categorya

 0–6 days, N = 476
 
7–27 days, N = 446
 
28–59 days, N = 160
 
 N (%) N (%) N (%) 
Prematurity, LBW 19 (4) (1) (1) 
Birth asphyxia (0) (0) (0) 
Birth injury 13 (3) 13 (3) (1) 
Congenital malformation 17 (4) 22 (5) 16 (10) 
Hyperbilirubinemia, physiological jaundice 264 (55) 223 (50) 11 (7) 
Sepsis 16 (3) (2) (1) 
Pneumonia, ALRI (0) (0) (1) 
URTI, mild ARI (0) 29 (7) 53 (33) 
Skin infection (0) (1) (1) 
Skin condition (non-infectious) (2) (2) 14 (9) 
Eye infection 25 (5) 44 (10) 17 (11) 
Oral thrush (0) (1) (1) 
Umbilical infection, umbilical problem (1) (2) (2) 
Diarrhea, dysentery, persistent diarrhea, dehydration 16 (3) (1) (5) 
Colic (1) 37 (8) 27 (17) 
Feeding problem 51 (11) 54 (12) (5) 
Gastrocolic reflex, regurgitation, constipation (0) (0) (1) 
Other 35 (7) 28 (6) 17 (11) 
 0–6 days, N = 476
 
7–27 days, N = 446
 
28–59 days, N = 160
 
 N (%) N (%) N (%) 
Prematurity, LBW 19 (4) (1) (1) 
Birth asphyxia (0) (0) (0) 
Birth injury 13 (3) 13 (3) (1) 
Congenital malformation 17 (4) 22 (5) 16 (10) 
Hyperbilirubinemia, physiological jaundice 264 (55) 223 (50) 11 (7) 
Sepsis 16 (3) (2) (1) 
Pneumonia, ALRI (0) (0) (1) 
URTI, mild ARI (0) 29 (7) 53 (33) 
Skin infection (0) (1) (1) 
Skin condition (non-infectious) (2) (2) 14 (9) 
Eye infection 25 (5) 44 (10) 17 (11) 
Oral thrush (0) (1) (1) 
Umbilical infection, umbilical problem (1) (2) (2) 
Diarrhea, dysentery, persistent diarrhea, dehydration 16 (3) (1) (5) 
Colic (1) 37 (8) 27 (17) 
Feeding problem 51 (11) 54 (12) (5) 
Gastrocolic reflex, regurgitation, constipation (0) (0) (1) 
Other 35 (7) 28 (6) 17 (11) 

ALRI, acute lower respiratory infection; URTI, upper respiratory tract infection.

aRaw frequencies with percentages of each group. (Percentages do not sum to 100 since other categories were omitted and infants may have zero or several secondary diagnoses).

Predictive value of symptoms and signs

The prevalence, sensitivity and specificity of individual signs for prediction of serious illness requiring urgent hospital management is presented in Table 5 by age group. In young neonates, sensitivity of individual clinical signs was above 35% only for measured temperature ≥37.5°C (65%). All signs had a specificity >85% and most values were >95%. ORs for association of individual clinical signs with need for urgent hospital management, considering signs with prevalence >2%, were highest and statistically significant for the following signs: history of difficulty feeding, not feeding well on feeding assessment and fever (temperature ≥37.5°C). Other promising signs for predicting serious illness requiring hospitalization (i.e. OR > 5) but too low in prevalence (<2%) for robust assessment of utility included movement only when stimulated, lethargy, hypothermia (temperature < 35.5°C), prolonged capillary refill, respiratory rate ≥60/min, history of convulsions, bulging fontanelle, history of blood in the stool, many or severe skin pustules and hardening of the skin.

Table 5

Prevalence, sensitivity, specificity and OR of individual signs for prediction of ‘serious illness requiring urgent hospital management’ (excluding jaundice), for infants

 Prev (%) Sens (%) Spec (%) OR (95% CI) 
Panel A: Infants aged 0–6 days, N = 476a      
General signs and symptoms 
History of fever 11.3 34.8 89.8 4.7 (1.9, 11.7) 
History of difficulty feeding 3.0 21.7 98.0 13.6 (4.1, 44.8) 
Not feeding well 4.1 21.7 96.8 8.4 (2.7, 25.8) 
Movement only when stimulated 1.1 8.7 99.3 14.3 (2.3, 90.1) 
Lethargic 0.8 8.7 99.6 21.5 (2.9, 160.0) 
Restless and irritable 0.6 0.0 99.3 2.7 (0.1, 54.6) 
Temp <35.5°C 0.8 13.0 99.8 67.8 (6.7, 681.0) 
Temp ≥37.5°C 14.9 65.2 87.6 13.3 (5.4, 32.8) 
Prolonged capillary refill 0.4 8.7 100.0 105.2 (4.9, 2260.3) 
Respiratory signs 
Respiratory rate ≥60/min 1.7 8.7 98.7 7.1 (1.4, 37.3) 
Neurological signs 
History of no cry at birth 8.8 4.5 90.9 0.5  
History of convulsions 0.2 0.0 99.8 6.4  
Bulging fontanelle 0.2 0.0 99.8 6.7  
Signs and symptoms related to diarrhea 
History of diarrhea 1.9 0.0 98.0 1.0 (0.1, 17.6) 
History of blood in the stool 0.2 0.0 99.8 6.4  
Reduced skin turgor 1.9 13.0 98.7 11.2 (2.6, 47.8) 
Other signs and symptoms 
History of feeding problem since birth 1.1 0.0 98.9 1.8 (0.1, 33.4) 
Many or severe skin pustules 0.4 4.3 99.8 20.5 (1.2, 338.7) 
Purulent drainage eyes 6.5 4.3 93.4 0.6 (0.1, 4.9) 
Redness around umbilicus 8.7 8.7 91.4 1.0 (0.2, 4.5) 
Pus from umbilicus 5.5 8.7 94.7 1.7 (0.4, 7.7) 
Hardening of skin 0.4 8.7 100.0 105.2 (4.9, 2260.3) 
Abdominal distension 11.9 0.0 87.6 0.2 (0.2, 13.3) 
Panel B: Infants aged 7–27 days, N = 446b      
Number of infants 446 17 429   
General signs and symptoms 
History of fever 15.5 41.2 85.5 4.1 (1.5, 11.3) 
History of difficulty feeding 6.7 23.5 93.9 4.8 (1.5, 15.7) 
Not feeding well 5.0 23.5 95.8 7.0 (2.1, 23.6) 
Movement only when stimulated 1.1 11.8 99.3 18.8 (2.9, 121.3) 
Lethargic 1.1 5.9 99.1 6.6 (0.7, 62.8) 
Restless and irritable 0.9 5.9 99.3 8.9 (0.9, 90.1) 
Temp. <35.5°C 0.5 5.9 99.8 26.8 (1.6, 447.2) 
Temp. ≥37.5°C 6.7 29.4 94.2 6.7 (2.2, 20.6) 
Respiratory signs 
Respiratory rate ≥60/min 3.6 5.9 96.5 1.7 (0.2, 13.9) 
Nasal flaring 0.0 0.0 100.0   
Grunting 0.2 5.9 100.0 78.1 (3.1, 1990.4) 
Cyanosis 0.2 0.0 99.8 8.2 (0.3, 207.6) 
Neurological signs 
History of no cry at birth 12.0 20.0 88.3 1.9 (0.5, 6.9) 
History of convulsions 0.2 6.3 100.0 82.7 (3.2, 2113.9) 
Signs and symptoms related to diarrhea 
History of diarrhea 4.9 5.9 95.1 1.2 (0.2, 9.6) 
Reduced skin turgor 1.1 11.8 99.3 18.9 (2.9, 121.6) 
Other signs and symptoms 
History of feeding problem since birth 2.0 0.0 97.9 1.3 (0.1, 23.9) 
Many or severe skin pustules 0.7 11.8 99.8 56.9 (4.9, 663.1) 
Purulent drainage eyes 14.6 11.8 85.3 0.8 (0.2, 3.5) 
Purulent drainage ears 0.2 0.0 99.8 8.1 (0.3, 207.1) 
Redness around umbilicus 6.3 5.9 93.7 0.9 (0.1, 7.3) 
Pus from umbilicus 7.9 11.8 92.3 1.6 (0.3, 7.3) 
Stiff limbs 0.2 0.0 99.8 8.2 (0.3, 207.6) 
Abdominal distension 11.9 0.0 87.6 0.2 (0.0, 3.4) 
Panel C: Infants aged 28–59 days, N = 160c 
Number of infantsa 314 37 277   
General signs and symptoms 
History of fever 33.1 75.0 67.9 6.4 (0.6, 62.7) 
History of difficulty feeding 1.9 25.0 98.7 25.7 (1.8, 366.4) 
Not feeding well 5.7 50.0 95.5 21.1 (2.6, 172.9) 
Movement only when stimulated 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Lethargic 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Restless and irritable 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Temp. <35.5°C 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Temp. ≥37.5°C 12.0 50.0 89.0 8.1 (1.1, 61.4) 
Respiratory signs 
Respiratory rate ≥60/min 5.0 25.0 95.5 7.1 (0.7, 77.2) 
Cyanosis 1.3 0.0 98.7 6.9 (0.3, 164.8) 
Neurological signs      
History of no cry at birth 18.3 50.0 82.6 4.7 (0.6, 35.1) 
Signs and symptoms related to diarrhea 
History of diarrhea 4.4 25.0 96.2 8.3 (0.8, 92.4) 
History of blood in the stool 1.3 25.0 99.4 51.7 (2.6, 1035.7) 
Sunken eyes 1.3 25.0 99.4 51.7 (2.6, 1035.7) 
Reduced skin turgor 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Other signs and symptoms 
History of feeding problem since birth 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Many or severe skin pustules 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Purulent drainage eyes 12.5 0.0 87.2 0.7 (0.0, 14.3) 
Purulent drainage ears 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Redness around umbilicus 0.6 0.0 99.4 11.4 (0.4, 321.4) 
Pus from umbilicus 1.3 0.0 98.7 6.9 (0.3, 164.8) 
Hardening of skin 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Abdominal distension 29.4 25.0 70.5 0.8 (0.1, 7.9) 
 Prev (%) Sens (%) Spec (%) OR (95% CI) 
Panel A: Infants aged 0–6 days, N = 476a      
General signs and symptoms 
History of fever 11.3 34.8 89.8 4.7 (1.9, 11.7) 
History of difficulty feeding 3.0 21.7 98.0 13.6 (4.1, 44.8) 
Not feeding well 4.1 21.7 96.8 8.4 (2.7, 25.8) 
Movement only when stimulated 1.1 8.7 99.3 14.3 (2.3, 90.1) 
Lethargic 0.8 8.7 99.6 21.5 (2.9, 160.0) 
Restless and irritable 0.6 0.0 99.3 2.7 (0.1, 54.6) 
Temp <35.5°C 0.8 13.0 99.8 67.8 (6.7, 681.0) 
Temp ≥37.5°C 14.9 65.2 87.6 13.3 (5.4, 32.8) 
Prolonged capillary refill 0.4 8.7 100.0 105.2 (4.9, 2260.3) 
Respiratory signs 
Respiratory rate ≥60/min 1.7 8.7 98.7 7.1 (1.4, 37.3) 
Neurological signs 
History of no cry at birth 8.8 4.5 90.9 0.5  
History of convulsions 0.2 0.0 99.8 6.4  
Bulging fontanelle 0.2 0.0 99.8 6.7  
Signs and symptoms related to diarrhea 
History of diarrhea 1.9 0.0 98.0 1.0 (0.1, 17.6) 
History of blood in the stool 0.2 0.0 99.8 6.4  
Reduced skin turgor 1.9 13.0 98.7 11.2 (2.6, 47.8) 
Other signs and symptoms 
History of feeding problem since birth 1.1 0.0 98.9 1.8 (0.1, 33.4) 
Many or severe skin pustules 0.4 4.3 99.8 20.5 (1.2, 338.7) 
Purulent drainage eyes 6.5 4.3 93.4 0.6 (0.1, 4.9) 
Redness around umbilicus 8.7 8.7 91.4 1.0 (0.2, 4.5) 
Pus from umbilicus 5.5 8.7 94.7 1.7 (0.4, 7.7) 
Hardening of skin 0.4 8.7 100.0 105.2 (4.9, 2260.3) 
Abdominal distension 11.9 0.0 87.6 0.2 (0.2, 13.3) 
Panel B: Infants aged 7–27 days, N = 446b      
Number of infants 446 17 429   
General signs and symptoms 
History of fever 15.5 41.2 85.5 4.1 (1.5, 11.3) 
History of difficulty feeding 6.7 23.5 93.9 4.8 (1.5, 15.7) 
Not feeding well 5.0 23.5 95.8 7.0 (2.1, 23.6) 
Movement only when stimulated 1.1 11.8 99.3 18.8 (2.9, 121.3) 
Lethargic 1.1 5.9 99.1 6.6 (0.7, 62.8) 
Restless and irritable 0.9 5.9 99.3 8.9 (0.9, 90.1) 
Temp. <35.5°C 0.5 5.9 99.8 26.8 (1.6, 447.2) 
Temp. ≥37.5°C 6.7 29.4 94.2 6.7 (2.2, 20.6) 
Respiratory signs 
Respiratory rate ≥60/min 3.6 5.9 96.5 1.7 (0.2, 13.9) 
Nasal flaring 0.0 0.0 100.0   
Grunting 0.2 5.9 100.0 78.1 (3.1, 1990.4) 
Cyanosis 0.2 0.0 99.8 8.2 (0.3, 207.6) 
Neurological signs 
History of no cry at birth 12.0 20.0 88.3 1.9 (0.5, 6.9) 
History of convulsions 0.2 6.3 100.0 82.7 (3.2, 2113.9) 
Signs and symptoms related to diarrhea 
History of diarrhea 4.9 5.9 95.1 1.2 (0.2, 9.6) 
Reduced skin turgor 1.1 11.8 99.3 18.9 (2.9, 121.6) 
Other signs and symptoms 
History of feeding problem since birth 2.0 0.0 97.9 1.3 (0.1, 23.9) 
Many or severe skin pustules 0.7 11.8 99.8 56.9 (4.9, 663.1) 
Purulent drainage eyes 14.6 11.8 85.3 0.8 (0.2, 3.5) 
Purulent drainage ears 0.2 0.0 99.8 8.1 (0.3, 207.1) 
Redness around umbilicus 6.3 5.9 93.7 0.9 (0.1, 7.3) 
Pus from umbilicus 7.9 11.8 92.3 1.6 (0.3, 7.3) 
Stiff limbs 0.2 0.0 99.8 8.2 (0.3, 207.6) 
Abdominal distension 11.9 0.0 87.6 0.2 (0.0, 3.4) 
Panel C: Infants aged 28–59 days, N = 160c 
Number of infantsa 314 37 277   
General signs and symptoms 
History of fever 33.1 75.0 67.9 6.4 (0.6, 62.7) 
History of difficulty feeding 1.9 25.0 98.7 25.7 (1.8, 366.4) 
Not feeding well 5.7 50.0 95.5 21.1 (2.6, 172.9) 
Movement only when stimulated 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Lethargic 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Restless and irritable 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Temp. <35.5°C 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Temp. ≥37.5°C 12.0 50.0 89.0 8.1 (1.1, 61.4) 
Respiratory signs 
Respiratory rate ≥60/min 5.0 25.0 95.5 7.1 (0.7, 77.2) 
Cyanosis 1.3 0.0 98.7 6.9 (0.3, 164.8) 
Neurological signs      
History of no cry at birth 18.3 50.0 82.6 4.7 (0.6, 35.1) 
Signs and symptoms related to diarrhea 
History of diarrhea 4.4 25.0 96.2 8.3 (0.8, 92.4) 
History of blood in the stool 1.3 25.0 99.4 51.7 (2.6, 1035.7) 
Sunken eyes 1.3 25.0 99.4 51.7 (2.6, 1035.7) 
Reduced skin turgor 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Other signs and symptoms 
History of feeding problem since birth 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Many or severe skin pustules 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Purulent drainage eyes 12.5 0.0 87.2 0.7 (0.0, 14.3) 
Purulent drainage ears 0.6 0.0 99.4 11.5 (0.4, 323.4) 
Redness around umbilicus 0.6 0.0 99.4 11.4 (0.4, 321.4) 
Pus from umbilicus 1.3 0.0 98.7 6.9 (0.3, 164.8) 
Hardening of skin 0.6 25.0 100.0 134.1 (4.6, 3906.2) 
Abdominal distension 29.4 25.0 70.5 0.8 (0.1, 7.9) 

aWe did not find: Nasal flaring, grunting, cyanosis, severe chest in drawing, diarrhea >14 days, sunken eyes, stiff limbs, purulent drainage from the ears.

bWe did not find: nasal flaring, severe chest in drawing, prolonged capillary refill, diarrhea >14 days, blood in the stool, sunken eyes, hardening of the skin.

cWe did not find: nasal flaring, grunting, prolonged capillary refill, severe chest in drawing, diarrhoea >14 days, convulsions, bulging fontanelle, stiff limbs.

Symptoms and clinical signs identified by person A had similar predictive value in the late neonatal period (Table 5, Panel B). Individual sensitivity was generally low, and specificity was high, as for the early neonatal period. Prevalence of signs was also similar to the early neonatal period, except that diarrhea and purulent eye drainage were seen more commonly, and delayed capillary refill was not found. The odds of hospital admission as a function of presence of symptoms and clinical signs also followed a similar pattern, except that history of difficulty feeding and respiratory rate >60/min were less predictive.

In the second month of life, history of diarrhea emerged as a useful sign, while not feeding well on assessment increased in utility.

Discussion

This study was designed to evaluate signs of severe illness in children <2 months of age, in the city of La Paz, where most of the population had good access to health care. It is necessary to identify signs of severe illness in young infants by health workers, so that patients can be promptly referred to health facilities and adequate care provided. Some signs have been introduced in the Bolivian national version of IMCI, and the IMCI adaptations of other countries in the America–Caribbean region, for children <7 days of age, and these signs need to be properly validated [14–17, 21–23, 26].

This study had the limitation that a large majority of patients had minor illness or were well on physical examination. Therefore, the number of sick children or babies needing referral for hospital care was small. This may reflect two primary factors: (i) the study took place at two tertiary care hospitals which provide primary healthcare for patients in the catchment areas, and the study by design excluded patients who were referred to the hospital by a health care provider in an attempt to capture a spectrum of patients that reflect those from the community who self-refer for care, and (ii) patients were part of a national insurance system (SUMI) for mothers and children <5 years of age, and thus there was no cost to patients to come to the hospital, even for minimal complaints and minor illnesses. Furthermore, nearly all patients (>90%) had a healthy start as they were delivered by a skilled attendant to mothers who had good access to antenatal care and were breastfed. Furthermore, fewer than 10% were preterm or had LBW. However, the babies in our series had lower than average z-scores for weight, length and head circumference, which may be explained in part by the high altitude [27].

The most common presenting complaint for young infants in this study was jaundice. Similarly, the most common diagnosis in the neonatal period was hyperbilirubinemia/physiological jaundice for both admitted and non-admitted patients. Feeding problem was also a relatively common secondary diagnosis, and is likely associated with jaundice as a result of dehydration. Correct breastfeeding techniques need to be further emphasized; perhaps improved orientation of mothers to breastfeeding could reduce this problem complex. Despite the importance of hyperbilirubinemia, the utility of clinical signs for first-line health workers to identify infants needing hospital management of this condition is not known [18]. Assessment of the predictive value of clinical signs of jaundice will be reported elsewhere as a combined analysis with the other sites in this multicenter study.

Most symptoms and clinical signs evaluated in this study were not good predictors of severe illness requiring management in the hospital in Bolivia. In general, signs performed similarly in the early and late neonatal periods and the second month of life. The most useful signs included history of difficulty feeding, not feeding well on feeding assessment and fever (temperature ≥37.5°C), signs which other studies suggest are useful for identifying sick young infants [18–22, 24, 28–31]. A number of other clinical signs showed predictive potential for identifying the sick young infant needing admission, however, their prevalence was too low in this study to draw firm conclusions.

Bolivia recently introduced an adaptation of IMCI which encompasses the young neonate in the first week of life. In general, this study supports the use of the clinical signs included in that algorithm for recognition of the sick neonate. Use of IMCI has significant potential for reducing mortality of young infants in Bolivia, although further monitoring and evaluation of the program is needed as it rolls out [32].

Funding

This study was supported by the Saving Newborn Lives program of Save the Children-US through a grant from the Bill & Melinda Gates Foundation.

Acknowledgements

We thank the directors of HN and HMI for their support of the study. We also thank the Bolivian Pediatric Society of La Paz, especially her President Dra. Luz Gabriela Valverde for their assistance with project management and review of patient case report forms. The Bolivian office of Save the Children-US provided administrative support. David Hamer at Boston University provided assistance with project oversight and made a site visit as a member, with Dr Gary L. Darmstadt and Dr Martin Weber, of the multicenter study Steering Committee. We thank Philip Greenwood and Claudine Chionh, Murdoch Children’s Research Institute, for data management and data processing.

References

1
Saving Newborn Lives
State of the World’s Newborns
 , 
2001
Washington, DC
Save the Children
(pg. 
1
-
41
)
2
Lawn
JE
Wilczynska-Ketende
K
Cousens
SN
Estimating the causes of 4 million neonatal deaths in the year 2000
Int J Epidemiol
 , 
2006
, vol. 
35
 (pg. 
706
-
18
)
3
Lawn
JE
Cousens
S
Zupan
J
4 million neonatal deaths: when? Where? Why?
Lancet
 , 
2005
, vol. 
365
 (pg. 
891
-
900
)
4
Gutiérrez
M
Ochoa
LH
Castillo
W
National Health and Demographics Survey 2003
 , 
2004
La Paz
Instituto Nacional de Estadística
5
Mazzi
E
Mansilla
G
En Mazzi
E
Sandoval
O
Neonatal IMCI
Perinatology
 , 
2002
2nd
La Paz
Elite Impresiones
(pg. 
310
-
3
)
6
Bang
AT
Bang
RA
Baitule
SB
, et al.  . 
Effect of home-based neonatal care and management of sepsis on neonatal mortality: field trial in rural India
Lancet
 , 
1999
, vol. 
354
 (pg. 
1955
-
61
)
7
Manandhar
DS
Osrin
D
Shrestha
BP
, et al.  . 
Effect of a participatory intervention with women’s groups on birth outcomes in Nepal: cluster-randomised controlled trial
Lancet
 , 
2004
, vol. 
364
 (pg. 
970
-
9
)
8
Jokhio
AH
Winter
HR
Cheng
KK
An intervention involving traditional birth attendants and perinatal and maternal mortality in Pakistan
N Engl J Med
 , 
2005
, vol. 
352
 (pg. 
2091
-
9
)
9
Bhutta
ZA
Darmstadt
GL
Hasan
B
, et al.  . 
Community-based interventions for improving perinatal and neonatal health outcomes in developing countries: review of the evidence
Pediatrics
 , 
2005
, vol. 
115
 (pg. 
519
-
617
)
10
Darmstadt
GL
Bhutta
ZA
Cousens
S
, et al.  . 
Evidence-based, cost-effective interventions: how many newborn babies can we save?
Lancet
 , 
2005
, vol. 
365
 (pg. 
977
-
88
)
11
Haws
RA
Thomas
AL
Bhutta
ZA
, et al.  . 
Impact of packaged interventions on neonatal health: a review of the evidence
Health Policy Plann
 , 
2007
, vol. 
22
 (pg. 
193
-
215
)
12
Darmstadt
GL
Walker
N
Lawn
JE
, et al.  . 
Saving newborn lives in Asia and Africa: cost and impact of phased scale-up of interventions within the continuum of care
Health Policy Plann
 , 
2008
, vol. 
23
 (pg. 
101
-
17
)
13
Baqui
AH
Arifeen
SE
Darmstadt
GL
, et al.  . 
Effect of a package of community-based newborn care delivered by two strategies in Sylhet district, Bangladesh: a cluster-randomised controlled trial
Lancet
 , 
2008
, vol. 
371
 (pg. 
1936
-
44
)
14
Mazzi
E
Cordero
D
Mejia
M
IMCI strategy in Bolivia
Rev Soc Bol Ped
 , 
1997
, vol. 
36
 (pg. 
106
-
8
)
15
Mazzi
E
What is your experience in relation to the Integrated Management of Childhood Illnesses (IMCI)?
Arch Argent Pediatr
 , 
1999
, vol. 
97
 (pg. 
408
-
10
)
16
Integrated Management of Childhood Illnesses (IMCI)
Health Ministry. General Health Services Agency
Children and Adolescent National Health Program
 , 
2003
La Paz
Ministerio de Salud y Deportes
17
Badner
A
Bartos
A
Cornejo
C
Delgadillo
C
Prieto
C
Espada
R
Clinical Neonatal IMCI. 3rd edn. Ministerio de Salud y Deportes
 , 
2005
La Paz
Editorial Jimenez
18
Darmstadt
GL
Baqui
AH
Choi
Y
, et al.  . 
Validation of community health workers’ assessment of neonatal illness in rural Bangladesh
Bull WHO
 , 
2009
, vol. 
87
 (pg. 
12
-
9
)
19
Kalter
HD
Burnham
G
Kolstad
PR
, et al.  . 
Evaluation of clinical signs to diagnose anaemia in Uganda and Bangladesh, in areas with and without malaria
Bull World Health Organ
 , 
1997
, vol. 
75
 
Suppl 1
(pg. 
103
-
11
)
20
Kalter
HD
Schillinger
JA
Hossain
M
, et al.  . 
Identifying sick children requiring referral to hospital in Bangladesh
Bull World Health Organ
 , 
1997
, vol. 
75
 
Suppl 1
(pg. 
65
-
75
)
21
Young Infants Study Group. The WHO
Clinical predictors of serious infections in young infants in developing countries
Pediatr Infect Dis J
 , 
1999
, vol. 
10
 
Suppl
(pg. 
S23
-
31
)
22
Weber
MW
Carlin
JB
Gatchalian
, et al.  . 
Predictors of neonatal sepsis in developing countries
Pediatr Infect Dis J
 , 
2003
, vol. 
22
 (pg. 
711
-
6
)
23
World Health Organization (WHO)
Integrated management of childhood illness
 , 
2002
Geneva
WHO
 
Available at URL: http://www.who.int/child-adolescent-health/integ/htm (9 June 2009, date last accessed)
24
Young Infants Clinical Signs Study Group
Clinical signs predicting severe illness in young infants: a multicentre study
Lancet
 , 
2008
, vol. 
371
 (pg. 
135
-
42
)
25
World Health Organization
Pocket Book of Hospital Care
Guidelines for the Management of Common Illnesses with Limited Resources
 , 
2004
Geneva, Switzerland
WHO
26
Hagan
L
Simple measures could save “millions” of newborns
BMJ
 , 
2001
, vol. 
323
 (pg. 
652
-
5
)
27
Mazzi
E
Díaz
M
Badner
A
, et al.  . 
Low birth weight infants
Risk factors. Rev Soc Bol Ped
 , 
1995
, vol. 
34
 (pg. 
83
-
7
)
28
Mullany
LC
Darmstadt
GL
Katz
J
, et al.  . 
Development of clinical sign based algorithms for community based assessment of omphalitis
Arch Dis Child Fetal Neonatal Ed
 , 
2006
, vol. 
91
 (pg. 
F99
-
104
)
29
Weber
MW
Mulholland
EK
Jaffar
S
, et al.  . 
Evaluation of an algorithm for the integrated management of childhood illness in an area with seasonal malaria in the Gambia
Bull World Health Organ
 , 
1997
, vol. 
75
 
Suppl 1
(pg. 
25
-
32
)
30
Bang
AT
Bang
RA
Reddy
MH
, et al.  . 
Simple clinical criteria to identify sepsis or pneumonia in neonates in the community needing treatment or referral
Pediatr Infect Dis J
 , 
2005
, vol. 
24
 (pg. 
335
-
41
)
31
Reddy
MH
Bang
AT
How to identify neonates at risk of death in rural India: clinical criteria for the risk approach
J Perinatol
 , 
2005
, vol. 
25
 
Suppl 1
(pg. 
S44
-
50
)
32
Bryce
J
Victora
CG
Habicht
JP
, et al.  . 
Programmatic pathways to child survival: results of a multi-country evaluation of Integrated Management of Childhood Illness
Health Policy Plan
 , 
2005
, vol. 
20
 
Suppl 1
(pg. 
i5
-
17
)

Appendix 1

Co-Principal Investigators: Eduardo Mazzi Gonzales de Prada, MD, Andres Bartos Miklos, MD; Study Physicians: Teresa Villagomez, Pablo Mattos, Manuel Pantoja Ludueña, Remedios Zumarán; Study Nurses: Irma Quispe, Willy Tarqui, Lourdes Checa and Claudia Canqui; Data Management: Erick Dueñas, Omar Vargas.

Author notes

Present address: Gary L. Darmstadt, Integrated Health Solutions Development, Global Health Program, Bill & Melinda Gates Foundation, BO Box 23350, Seattle, WA 98102, USA.
The Members of the Bolivia Clinical Signs Study Group are listed in the Appendix 1.

Comments

0 Comments