Report on the 10th Annual College of Medicine Research Dissemination Meeting

Dr SM Graham

The reagent costs per BlantyreCount assay were US$0.44 for both absolute count and %CD4/lymphocyte, and US$0.11 for %CD4/lymphocyte alone. The 95% limits of agreement for BlantyreCount and TruCount for absolute counts in the CD4 range <400 cells/□l were −48.9 to 27.0 cells/□l and for %CD4/lymphocyte were −2•42% to 2•37%. Using BlantyreCount in the ART clinic in Blantyre, 37% of 143 patients with clinical stage I or II disease who were ineligible for ART by clinical staging criteria, had a CD4 count <200 cells/□l. In the same group, 47% of 34 patients with stage III disease, had CD4 counts >200 cells/□l.

Conclusion And Recommendations

BlantyreCount is a low-cost accurate flow cytometric method of CD4 counting which could improve clinical decision making about when to commence ART.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/05 - Anthropological And Bioethics Study of Clinical Research in Malawi

JM Mfutso-Bengo ¹, M Mkunthi ¹, V Jumbe ¹, E Robson ¹, F Masiye ¹, M Molyneux ¹

Abstract

Introduction

This research project is being conducted in three phases over three years. The first phase has been completed while Phase 2 is just about to begin; and Phase 3 will commence after completion of Phase 2. This presentation is based on the first phase only which is an Anthropological Study.

Objectives

To provide a base for informing, reforming and improving informed consent policy and practice by describing the local cultural attitudes and perceptions to research, autonomy, informed consent process and community consultation.

Setting

The rural areas surrounding Mpemba and Madziabango Health Centres and the urban areas surrounding Bangwe Health Centre in Blantyre District.

Methodology

Data was collected from a total of 494 participants between February and April 2005 using focus group discussions (FGDs). A total of 50 FGDs composed of 6–12 people per group were conducted in the three areas altogether. In an attempt to achieve homogeneity in the groups, the respondents were divided into the five categories of participants, non-participants, refusers, local leaders, and health workers. Sex and age composition varied greatly among the groups. Data analysis was done using N6 (also known as NUD*IST) with pre-categorized themes of understanding of health research, motivation to take part in research, perception of the informed consent process (individual consent and community consultation), and attitude towards health research.

Results

Apart from the category of health workers, the perception of the meaning of health research was similar among all respondents. Overall, the respondents were unable to accurately define health research and most commonly described health research as activities associated with preventive health measures such as community health assessment and education. They also differentiated health research from normal health care in terms of quality of care; thus normal health care in health centres was perceived as being unsatisfactory with medication given without proper diagnosis, and patients accorded poor reception; whereas in health research treatment is prompt and given after a proper diagnosis.

Many focus group participants who had participated or were participating in health research and non-participants cited wanting to know one's health status as a motivating factor to take part in research. Although this was overwhelmingly cited as a motivating factor, it also reinforces the idea that many people are not able to differentiate health research from seeking normal health care (therapeutic misconception). On the other hand, those who had refused to take part cited fear as the main reason because there are so many rumors that circulate in the communities once a research projects is launched. For example is research studies that involve drawing of blood samples the rumors are that the blood will be sold or will be used for satanic rituals.

All respondents strongly believed that they have to be asked to give their consent before being recruited in any research. The majority of participants said they preferred signing or thumb printing as opposed to consenting verbally. The major reason cited in preference of this method was that the signature can be used as a reference/evidence in future if something goes wrong. Very few of the focus group participants expressed a preference for verbal consent as it was perceived as faster than signing and because signing induces fear as a signature is seen as indelible and hence involves a measure of personal sacrifice.

The majority of the participants perceived community consultation to supersede individual consent. Before any health research starts, whether hospital or community based, FGD participants considered community consultation, through the local chief, to be a prerequisite. In their universal opinion this has to be done on the basis that it is customary for any outsider to first seek the consent of the chief, as a custodian of the land and subjects, before doing anything in the village.

Respondents expressed interest in health research projects to be conducted in their communities; but raised concerns that they are usually put off because many researchers do not seek prior consent from the community and people just hear about the research when they visit the clinics which makes them reluctant to take part as they are not sure what it is all about. They also expressed concern that researchers disappear after finishing conducting the research; and they do not come back to give feedback to the community which makes the participants lose interest.

Those in the category of “refusers” indicated that they would be interested to take part in research if the research procedures are clearly explained to them in order to array rumors that usually circulate such as the researchers collecting blood for sale or satanic rituals.

Conclusions

Though there is an impaired understanding of what health research means among the community members, the study established that people do not participate in health research because they think that the research is aimed at treating them; rather they choose to participate in health research for the sake of obtaining better quality treatment through the research procedures. While they can differentiate between routine health care and health research, the people actually make rational decisions to participate or refuse to participate in health research.

Contrary to the global thinking that those who refuse to take part in health research are emancipated and educated individuals in resource poor setting, it was found that the reasons for refusing were not based on knowledge; rather it was ignorance and superstition. Many believed that the researchers were drawing blood not for research, but for satanic rituals or for sale.

The study has also revealed that community members fully understand why they have to give consent prior to taking part in health research and many are aware that taking part in health research is voluntary. The majority preferred signing or printing their thumbs to giving verbal consent.

It has also been established that community consultation should be done before individuals from the communities are asked to take part in health research whether at the hospital or within the community.

Attitude towards health research is generally good among the people only that there are certain concerns that need to be addressed prior to recruitment of research participants. Things like adequate disclosure of research information and community consultation are of paramount importance to the people.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/06 - Anthropometric Characteristics of Underfive Children and Adults in Lungwena, Mangochi

DM Chilima ¹, KM Maleta ²

Abstract

Objective

To determine nutritional status of underfive children and adults in Lungwena area as a benchmark for evaluating the impact of the Lungwena Health, Nutrition and Agricultural Multidisciplinary Project.

Setting

Lungwena, Mangochi

Material and Methods

A baseline survey for the Lungwena multidisciplinary project was conducted from October to December 2004. A total of 300 households were randomly selected from the participating villages and a DHS format questionnaire was adopted for the survey. Anthropometric measurements such as weight, height and MUAC were taken among underfive children and adults (aged 15 –54 years for men and 12 – 49 years for women). To assess nutritional status, weight for age, height for age and weight for height were used for underfive children and body mass index (BMI) was used for adults.

Results

Overall, among underfive children, 36.4% were classified as stunted, 20% were underweight and 3.3% were wasted. Among adults, as expected, men were significantly taller (164.7±6.8 cm vs 153.6±7.4 cm, p <0.001) and heavier than women (56.0± 7.7 kg vs 49.3± 8.7 kg, p<0.001) but no difference was seen in MUAC. Overall one in ten men (12.2%) and two in ten women (20.8%) were malnourished using BMI <18.5kg/m2.

Conclusion And Recommendations

The baseline study has shown that malnutrition is a problem for both children and adults in the area. The challenge for the project is to improve the situation since the consequences of malnutrition can be severe.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/07 - Antibiotics for Prevention of Preterm Labour (Apple) Trial: A Randomised Controlled Trial

N van den Broek ^1,³, C Ntonya ¹, E Kayira ¹, G Kafulafula ², M Goodall ¹, S White ^1,⁴, JP Neilson ^1,⁵

Abstract

Objectives

We have previously described high rates of preterm birth in ultrasound-dated pregnant populations in Namitambo. There is a well established association of infections with preterm birth, and high burden of infection in this population. The objective of the APPLe study was to assess the effect of prophylactic azithromycin on various maternal and infant outcomes, primarily preterm birth.

Materials And Methods

Women attending for antenatal care in Bangwe, Mikolongwe, Namitambo and Nguludi, with gestational age less than 24 completed weeks (by ultrasound) at recruitment. This randomized controlled trial was of Azithromycin 2g orally, or placebo, given at <24 and 28–32 weeks. Ante- and post- natal care were provided in accordance with the Malawi Ministry of Health Program.

Results

1,149 women received Azithromycin; 1,148 placebo. Azithromycin did not reduce the incidence of preterm birth (< 37 weeks) which was 16.8% for the Azithromycin group and 17.4% for the placebo group. Risk factors for preterm delivery in this population will be reported as will the effects on secondary outcomes including birthweight and neonatal survival.

Conclusions And Recommendations

Although there was no demonstrable impact of Azithromycin on preterm birth, possible causes of preterm delivery deserve further exploration. Other factors that may contribute to preterm birth in this population may include infection, heavy work, stress and depression. A follow-up study of preterm, surviving babies is ongoing - to assess long term effects.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/08 - Antibody-dependent Complement-mediated Lysis: A Neglected Mechanism of Immunity To Salmonella

CA MacLennan ^1,^2,^3,⁴, E Gondwe ^1,⁵, C Msefula ^1,^2,⁵, L Wilson ¹, S White ^1,⁶, CA Hart ³, SM Graham ^1,⁵, ME Molyneux ^1,⁵, MT Drayson ⁴

Abstract

Objectives

To investigate the role of antibody and complement in humoral immunity to salmonella in Malawi.

Setting

MLW Research Programme and QECH.

Materials And Methods

Serum was collected from 65 medically-well children aged 0 to 10 years recruited from the surgical outpatient clinics at QECH and Cure Hospital. Serum killing of S. typhimurium strain D23580, isolated from a bacteraemic child at QECH, was assessed by serum bactericidal assay. Specific IgG and IgM to D23580 and complement-deposition on D23580 were measured by flow-cytometry.

Results

Serum from all children over 14 months of age effected a 1.4 log kill on D23580 (‘normal killing’). Of those under 14 months, 72% could not cause normal killing. All sera with specific IgG to D23580 were able to effect normal killing. 67% (8/12) sera that showed normal killing of D23580, but did not have specific IgG, were found to have specific IgM for D23580. Deposition of complement on D23580 was associated with killing. All sera that could deposit the membrane attack complex of complement on the surface of D23580 could effect normal killing. The amount of membrane attack complex deposited on D23580 showed strong correlation with specific IgG titre to D23580 (rs= 0.82, P<0.001).

Conclusion And Recommendations

These findings suggest that antibody-dependent complement mediated lysis is an important mechanism for immunity against salmonella in Malawi. The lack of specific antibody to salmonella in children under two years may explain why salmonella bacteraemia is particularly common in this age group. These findings support the development of a vaccine capable of inducing protective antibody in young Malawian children.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/09 - Arv Therapy Clinic, Queen Elizabeth Central Hospital, Blantyre: Retrospective Description Of Defaulters

TK Hartung ¹, G Nyirongo ², JJG van Oosterhout ¹, HW Hofland ¹, EE Zijlstra ¹

Abstract

Objective

To describe patients who have defaulted free-of-charge antiretroviral therapy (ART) between July 2004 and December 2005.

Methods

In June 2006 all ART files were reviewed retrospectively. Defaulting was defined as non-attendance for 3 consecutive months. Where appropriate, data were compared with cumulative national results.

Results

Of 3110 patients ever started over 18 months, 346 (11.1%; national 9%) defaulted after a mean of 4.5 months (sd 3.7) and 77% did so within the first 6 months. There was no difference between males and females. 33.5% had been transferred in mainly from private clinics at the launch of the ART program. Compared to the national cohort defaulters were more often in WHO stage IV (30% vs 24%, OR 1.5, 95% CI 1.17–1.91). During a total of 1352 follow-up visits mean body weight of defaulters did not rise, non-adherence as per pill count was 7.2% (97) and side effects occurred in 8.6% (116) of visits. There was no trace of the patients' whereabouts in 83% of cases but 10% were reported unwell at home and 7% had been admitted to hospital.

Conclusion And Recommendation

The defaulter rate was in line with national average. Although there was no active tracing of defaulters, we speculate that some may have died, in particular patients in WHO stage IV or those who were unwell at home or hospitalized. Others may have opted to return to private clinics after free ART became available for the private sector. Paying particular attention to patients who fail to gain weight might prevent defaulting.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/10 - Caesarean Sections and other Emergency Obstetric Surgery by Clinical Officers, Medical Officers and Specialists: The Malawi Experience

FW Kamwendo ¹, GC Chilopola ¹, E Malunga ¹, AM Chimbiri ², S Bergstrom ³, JG Malewezi ⁴, C Pereira ⁵, A Cumbi ⁵

Abstract

Objectives

19 (76%) of the 25 children below 3 years of age were described as non-active before starting ART compared with 1 child after taking ART for six months or longer. Severe motor developmental delay was found in 60% (n=15) of the children in this age group. 2 were classified as moderately and 3 as not delayed. On ART 64% (n=16) showed no delay for the assessed motor skills, 3 were still found to have severe and 1 infant had moderate developmental delay.

Conclusion

With the current national ART regimen not only the quantitative outcome but also the motor development and quality of life improve markedly. For the group of children between 3–5 years numbers were low and the assessment tool not specific enough to provide relevant data.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/20 - Increasing Uptake for HIV Testing and Counseling (HTC) Through a National Promotional Campaign

P Moses ¹, E Odoyo-June ³, F Kayambo ², E Libamba ¹, M Nyangulu ¹, C Sambakunsi ¹, J Mphaya ⁴, B Mhango ⁵, D Namate ¹

Abstract

Setting

Community mobilization for increased HTC uptake is critical to achievement of universal access targets for HIV prevention and care. Malawi therefore conducted the first National HTC week in July 2006.

Objective

To Conduct National HTC week targeting 50,000 clients and assess its success in increasing HTC uptake

Methods

Working with multisectoral partners, Ministry of Health and National AIDS Commission provided leadership in organizing and coordinating the HTC week. District level task forces developed and implemented plans for community mobilization and increasing local availability of HTC through outreaches. The National testing algorithm and standards were used.

Results

The 96,849 clients tested represented a ten-fold increase in average weekly HTC uptake. Over 70% of the clients were aged 21–34 years, children <15 years were 3%. About 11% (10,421) of clients tested HIV+. 53% of clients tested were female, having higher HIV+ rates than males (11.5% vs 8.2%; p< 0.0001) in all age-groups. Peak prevalence in females was higher and earlier (18.4% in 25–34 year age-group) than for male (12.2% in 35years+). HIV rate in children <15 years was double that for ages 15–24 years (11.8% vs 5.6%). Among clients below 15 years, females had higher HIV rate than males (14.0% vs 8.1%), suggesting a predominance of sexual transmission.

Conclusion and Recommendation

Increased availability of HTC services during campaign week effectively raised HTC uptake, and the initiative generated data for evaluating HTC program responsiveness. HIV prevalence trends seen in population data were further supported. Rapid equitable expansion of HTC services is recommended.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/21 - Infant Feeding Practices in a Rural and Semi-urban Community in Mangochi District

P Kamudoni ¹, G Holmboe-Ottesen ², K Maleta ³, Z Shi ⁴

Abstract

Objectives

To investigate socio-demographic factors associated with optimal infant feeding practices and nutrition status.

Methods

A cross sectional survey was conducted in 2004 where questionnaire data and infant's anthropometric measurements were obtained from 157 rural and 192 semi-urban infant - mother pairs.

Results

Early breastfeeding (<1hour after delivery) was practiced among 68.2% of the rural and 63% of the semi-urban mothers. Colostrum was given by 96%, while pre-lacteal feeds were given by 4%, of all the sampled mothers. Exclusive breastfeeding at 4 months was significantly higher (p<0.01) among semi-urban (44.4%) than rural (4.7%) mothers. 65.6% of the rural and 68% of the semi-urban mothers had introduced complementary feeds because of infant's crying. Living in the semi-urban area (OR = 7.94; 95% CI 2.03 – 31.1) and giving birth in a health facility (OR = 3.89; 95% CI 1.24 – 12.25) were independently associated with exclusive breastfeeding at 4 months.

12.6% of all the sampled infants were stunted and underweight, while 6.6% were wasted. Rates of stunting, being underweight or wasting were significantly higher (p<0.01) in the rural than semi-urban group; and among infants not exclusively breastfed at 4 months than those exclusively breastfed.

Conclusion and Recommendation

Semi-urban mothers are more likely to practice exclusive breastfeeding than rural mothers. Access to healthcare at birth is likely to influence mothers to exclusively breastfeed. Promotion of exclusive breastfeeding could be an effective way of addressing malnutrition in infancy. Provision of breastfeeding support through community health workers can be a way of promoting exclusive breastfeeding.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/22 - Influences on Decision-making for Pregnancy and Childbirth in a Rural Malawian Matrilineal Community

GD Barber ¹

Abstract

Objective

To examine relationships, influences and power dynamics underpinning childbearing decisions in a resource-poor subsistence farming community, southern Malawi.

Methods

Participant observation, key informant and group interviews of community residents, service providers and project workers, and household survey during nine months community residence. Childbearing themes, residence and relatedness, knowledge and power, health and risk perceptions were examined.

Audio and observational data was examined thematically, credibility being checked through triangulation and expert secondary translation of interpretation. Technical and anthropological literature was examined. Ethical issues were analysed.

Results

A strongly matrilineal and matrilocal lifestyle emerged, substantial decision-making power residing with pregnant women's older female relatives. Husbands have limited influence but are expected to support childbearing decisions made by older women and take responsibility for their own matrikin. Most young women and men remain too ignorant for independent action and formal health services do not necessarily take this into account.

Biomedical knowledge is layered over ethnomedical, and expressions of morality and wellbeing. Choices are ultimately pragmatic, neither biomedical nor ethnomedical prevailing. Nevertheless all age groups demonstrated acceptance of biomedical knowledge and service provision, and desired greater understanding.

Conclusions

The strong influence of older female matrikin has a substantial impact on childbearing choices so their willingness to engage positively with health services needs to be taken into account in maternity care provision. Excluding these decision-makers from encounters with pregnant women is counterproductive.

Recommendations

Action research and policy implementation that targets these matrikin, as well as men, to increase their knowledge and involvement may bring positive benefits for safer motherhood.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/23 - Insight, Psychopathology and Psychosocial Outcome in Schizophrenia

A Sefasi ¹, P Samalani ¹, N Crumlish ^1,², A Clohessy ¹, H Chilale ¹, E O'Callaghan ²

Abstract

Objective

To identify the determinants of psychosocial outcomes -functioning and quality of life -in a group of patients recruited to a randomised controlled trial (RCT) of carer education in schizophrenia.

Materials And Methods

Following diagnostic interview and recruitment to the RCT, we measured psychopathology, quality of life and functioning with standardised clinical measures. To evaluate the three dimensions of insight in psychosis - treatment compliance, recognition of illness and ability to relabel symptoms -we used the Schedule for Assessment of Insight.

Results

The best predictor of functioning was the ability to relabel symptoms (R2 change=0.22, p=0.001); predictors of poor functioning were negative symptoms (R2 change=0.16, p=0.001) and first-rank symptoms (R2 change=0.12, p=0.002). Poor quality of life was predicted by younger age (R2 change=0.13, p=0.02), hallucinations (R2 change=0.15, p=0.005), and increasing recognition of mental illness (R2 change=0.07, p=0.05).

Conclusions And Recommendations

These preliminary findings suggest that insight and psychopathology are important determinants of psychosocial outcomes in this group, pre-RCT. Perhaps surprisingly, different dimensions of insight had opposing effects on outcome, with recognition of mental illness showing an inverse relationship with quality of life. This may have implications for psychoeducation and disclosure of diagnosis in schizophrenia.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/24 - Lobar Pneumonia due to Nontyphoidal Salmonella in a Child with recent severe Malaria

L Mankhambo ¹, K Chiwaya ¹, A Phiri ¹, SM Graham ¹

To describe the nutritional and clinical outcomes of previously severely malnourished children commencing HAART.

Setting

Nutrition Rehabilitation Unit and Paediatric ARV clinic, QECH

Methods

Retrospective cohort study of Malawian children commenced on HAART following ARV scale up in September 2004. Age and gender adjusted height and weight z scores (WAZ, HAZ) and clinical outcome were studied for 93 severely malnourished children prior to HAART. Weight gain was analysed every 4 weeks to 6 months, and then at one year. Change in weight and height z score from starting HAART to last visit (median 36 weeks) were calculated. Defaulters with addresses were followed up.

Results

There was a significant improvement in mean WAZ score (− 4.04 to −2.63) after 26 weeks on HAART (n = 45, p=<0.01). There was however no significant improvement in stature by last visit (HAZ −4.05 to −3.86, n= 48, p= 0.226). The period of maximal weight gain was between the 10-–4th week after commencing HAART, and was 2.5g/Kg/d by 6 weeks without formal additional nutrition support. Of children commencing HAART, 73.6% were still attending clinic, with 16% known to have died, and 11.7% defaulted (outcome unknown). Weight loss at last clinic visit predicted death (p<0.01).

Conclusions And Recommendations

We report rapid improvement in weight, but not stunting in severely malnourished HIV infected children soon after starting HAART. Weight also appears a sensitive marker of failure of treatment, and should be routinely measured.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/28 - Outcome, HIV Serostatus and Testing Behaviour of Parents of HIV Infected Children at the Paediatric ART Clinic, Blantyre, Malawi

G Poerksen ¹, L Pollock ¹, F Bell ¹, P Tembo ¹, H Blencowe ¹, M Thindwa ¹, K Chiwaya ¹, E Chesshyre ¹, J Kenny ¹, O Jefferies ¹, E Molyneux ¹

Abstract

Objectives

To describe some features of the caretakers of HIV infected children from the paediatric ARV Clinic, QECH, Blantyre focusing on HIV sero-status, testing behaviour and percentage of guardians on antiretroviral treatment (ART).

Methods

498 children were registered in the paediatric ARV Clinic at the beginning of September 2006. Information about the relationship of the caretaker to the child, the outcome of the parents, their HIV sero-status and the number of parents on ART was collected from the file (30 of the original files were missing).

Results

The mother was the caretaker in 46.2% (n=230), followed by an aunt (n=73, 14.7%) and a grandmother (n=56, 11.2%). 70 children (14%) had lost their father, 78 (15.6%) their mother and 81 children both parents (16.2%). In 60 cases the outcome of the father was not definitely known due to separation or loss of contact after the death of the mother. Lifetable analysis showed a trend towards fewer deaths and defaulters in the group of orphaned or single parent children but did not differ significantly from the group of children who still had both of their parents. Of the mothers that were alive 157 (51.5%) reported to be tested HIV positive, of whom 61 (38.8%) stated to be taking ART. The fathers were reported to be positive in 21.1% (n=54) and almost half of them (n=27) were on ART. 33.3% of the mothers and 52.3% of the fathers said that they had not been tested. The HIV status was not recorded by the Clinician in about a third of the cases.

Conclusion

Results And Progress to Date

A total of 308 cancer patients were recruited, and 243 were newly diagnosed. A total of 215 mothers were interviewed. The five common tumours were Burkitt lymphoma at 50%, 10% Wilms tumour, 6% Rhabdomyosarcoma, 7% Kaposi sarcoma, and 6% Retinoblastoma. Eleven percent of the children and 20% of the mums were HIV positive, and syphilis was positive in 14% of mothers. Morphological verification was attained for 67% of the tumours and 30% were clinically or radiologically diagnosed. The data is still being thoroughly analysed, and blood samples being analysed for malaria antibodies and herpes viruses. Updated results will be presented.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/32 - Prevalence and Causes of Childhood Blindness in Southern Region of Malawi

K Kalua ¹, S Kanjoloti ²

Abstract

Objectives

To determine the prevalence and causes of childhood blindness in Chikwawa District, Malawi
To use the key informant method in identifying blind children in Chikwawa District, Malawi

Materials And Methods

This was a population based cross sectional survey of a catchment area comprising of 43,000 children in Chikwawa district. Key informants from various communities were identified through local village leaders and trained to identify blind and severe visual impaired children from the communities. Identified children had a proper eye examination at designated centres. Children requiring further medical assistance were referred to tertiary centres.

Results

Results were analysed using Stata software 9.0 and epi info version 6.4. 44 key informants were trained to cover 196 villages. In total 37 blind children were identified from the study area. The prevalence of childhood blindness was 0.09% (CI 0.06–0.11). Congenital cataract was the commonest cause of blindness (35%) followed by corneal disease (22%).

Conclusions And Recommendations

Key informant method is an easy way of identifying blind children in the community for purposes of programme intervention. The fact that prevalence of childhood blindness in Chikwawa (0.09%) obtained using the Key informant method is well within the WHO projected figures of blindness in a developing country indicates reliability and calls for recommendation of this method for other diseases.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/33 - Randomized Trial of Intermittent Treatment of Bacterial Vaginosis (BV) using Intravaginal Antibiotic Microbicide

TE Taha ¹, G Kafulafula ², B Makanani ², S Chen ¹, C Nkhoma ³, D Hoover ⁴, NI Kumwenda ³

Abstract

Background

We presumptively treated BV and genitourinary symptoms (GUS) using an intravaginal antibiotic microbicide.

Methods

A randomized, double-masked, placebo-controlled, phase III clinical trial of intermittent BV treatment was conducted in Malawi. Starting at enrollment and repeated every 3 months for one year, HIV-uninfected and infected women used metronidazole 0.75% gel (MetroGel-Vaginal®) or placebo gel intravaginally for 5 consecutive nights. At each visit, tests for BV, T. vaginalis and candida were performed and information on GUS was collected. Cross-sectional proportions of women with BV or GUS were compared using exact tests. Robust covariance log-binomial models assessed longitudinal associations of BV with treatment and other covariates.

Results

842 HIV-uninfected and 844 infected women were randomized equally to treatment or placebo. During the study, BV prevalence among HIV-uninfected women declined from 45.9% to 24.7% with treatment (p<0.001) and from 46.8% to 28.8% with placebo (p<0.001). Among HIV infected women there were also significant declines in BV prevalence over time. The metronidazole treatment advantage (vs placebo) persisted after adjusting for covariates (HIV-uninfected: adjusted relative risk [ARR] 0.88, 95% CI 0.81–0.96; HIV-infected ARR 0.94. 95% CI 0.89–1.00). Both treatment and placebo gels substantially reduced GUS frequency (>60%).

Conclusion

Intermittent use of vaginal gels should be considered to prevent vaginal infections.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/34 - Reproductive Intentions and Sexual Behaviors among Women aged 20–49 in Lungwena

G Kamvazina ¹, KM Maleta ², AM Chimbiri ¹

Abstract

Objectives

To assess women's reproductive intentions in avoidance of unintended pregnancy, establish sexual behaviors contributing to unintended pregnancies, and recommend strategies for strengthening sexual behavior change activities in Lungwena.

Setting

Six villages in TA Makanjira and Chowe in Lungwena health centre catchment in Mangochi district

Materials And Methods

Data were collected in a cross sectional study (Demographic and Health Survey {DHS}) conducted in 2004 by Lungwena NUFU project. Stratified random sampling was used to select 300 households from six villages in the catchment area. Enumerators administered adult and adolescent women's questionnaires in one-on-one interviews in Chiyao.

results

188 women aged 20–49 were interviewed. 35.5% and 30% of those who had unintended pregnancy at conception for first born and second born respectively, could have liked to wait for over 2 years. At the time of study, 65.5% wanted another child after 2 years. Among those who had sexual exposure one month prior to the survey, 62.6% had unprotected sex, >50% of all women wanted family size > 7 compared to 29.5% that had achieved this. Besides, only 27.1% perceived pregnancy in the next few weeks as a big problem, and spousal communication about family planning was low (22%). There was a discrepancy between reproductive intentions and sexual and reproductive behaviors.

Conclusions And Recommendations

Reproductive intentions among the women had no influence on sexual and reproductive behaviors. In this case, it is recommended to design behavior change interventions that address the specific behaviors for reduction of unintended pregnancies in Lungwena.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/35 - Return of Chloroquine Antimalarial Efficacy in Malawi

MK Laufer ¹, PC Thesing ², ND Eddington ¹, RS Masonga ¹, FK Dzinjalamala ¹, SL Takala ¹, TE Taylor ¹, CV Plowe ²

Abstract

Objectives

Abstract

Objectives

To explore accuracy and cost associated with using a serial rapid testing algorithm as an alternative to the standard parallel algorithm for HIV diagnosis.

Methods

Test results from all clients served from Jan-June 2006 at MACRO VCT in Lilongwe were reviewed to assess the potential impact of serial testing on the number of false results given to clients and cost of HIV testing.

Results

Of the 6810 client results included in this assessment, individual outcomes returned to the client under both parallel and serial testing algorithms would be identical. Under the parallel testing algorithm 99.3% of the outcomes required two tests and the remaining required three tests, at an estimated cost of US$14,036.50 ($2.06/ test result). However, if a serial testing algorithm had been used, then 84.8% of the samples would only require one test, 14.6% two tests, and 0.06% three tests, resulting in an overall cost of US$6865.90 ($1.01/ test result).

Conclusions

When comparing the observed results under parallel testing to those that would have been observed under serial testing in the MACRO data, we conclude that there is no detectable difference in accuracy. However, serial testing requires far fewer tests than parallel testing at half of the cost.

Recommendations

Based on this investigation, we recommend that a field assessment and a subsequent policy level discussion be held to assess the appropriateness of a serial testing algorithm for HIV diagnosis, with the potential implication of less costly HIV testing services.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/39 - Prospective Study of Aetiology of Severe Pneumonia in Children

A Phiri ¹, L Mankhambo ¹, K Chiwaya ¹, S Kaunda ¹, T Chikaonda ¹, EM Molyneux ¹, ME Molyneux ¹, SM Graham ¹

Abstract

Background

Pneumonia is a very common cause of admission to the paediatric wards at QECH and mortality is high. Failure to respond to currently recommended first-line antibiotics is very common.

Objectives

To describe the aetiology and outcome of HIV-infected and uninfected Malawian children 2 months and older admitted to the paediatric wards at QECH with a diagnosis of severe pneumonia as defined by WHO criteria.

Methods

A prospective study over a 12-month period is ongoing and will be completed by October 2006. Following informed consent, clinical and demographic data are collected on admission. Investigations for cause of pneumonia include blood culture, nasopharyngeal aspirate (NPA) and/or induced sputum samples for respiratory virus screen (RSV, adenovirus, influenza, parainfluenza), Pneumocystis and M.tuberculosis (including culture). Lung aspirates for bacterial culture are performed before antibiotic administration in a subgroup of children. An appropriate HIV test for age is done following separate consent. Penicillin and gentamicin are the first-line antibiotics and oxygen use is guided by pulse oximetry. Outcome measures include death, days in hospital and days requiring oxygen.

Results

Data are available for the first 200 children enrolled with severe pneumonia. Causes of pneumonia include bacteria (31 cases with Strep pneumoniae [n=11] and nontyphoidal Salmonella [n=9] the commonest with 4 each of Staph aureus and Haemophilus influenzae), Pneumocystis (11 cases) and tuberculosis (8 cases). No results of viral screen available. HIV infection status is known for 157 cases and HIV prevalence was 51%. Case-fatality rate was 5% overall but 64% for PcP.

Conclusion

It is difficult to confirm aetiology in most cases of severe pneumonia in children. The finding that nontyphoidal Salmonella is a common cause of severe bacterial pneumonia raises concerns about appropriate first-line antibiotic treatment. The case-fatality rate for this study group was much lower than for a similar study in 1996 (22%) and important differences in severity and HIV-related management will be analysed.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/40 - Social Support for Women Caregivers of Family Members with HIV and AIDS in Rural and Peri-urban in Malawi

G Msiska ¹, MC Nyando ¹, CN Chihana ¹, CPN Kaponda ¹

Abstract

Objectives

To explore social support for women caregivers in rural and peri-urban communities of Lilongwe

Setting

Lumbadzi and Mitundu, where Lumbadzi is peri-urban and Mitundu is predominantly rural

Methods

Social support is an important buffer for family caregivers of people living with HIV/AIDS. In order to further our understanding, an exploratory and descriptive qualitative study was conducted to explore social support for women caregivers in rural and urban communities of Lilongwe. Participants for the study involved women caregivers over the age of 18 years who were providing care to sick family members who have either HIV/AIDS or show symptoms that suggest HIV/AIDS. 4 focus groups, (8 participants in each) were conducted in setting, making a total of 8 focus groups. Focus group discussions were tape recorded and thereafter transcribed for analysis through which major themes were identified which included types of support caregivers receive and sources of support.

Results

Findings revealed that the support caregivers receive is quite substantial. The following categories of support were identified: physical, material, informational, emotional, financial and spiritual support. Material, financial and spiritual support were the ones commonly reported. However, some participants indicated that physical support is rarely given and others did not perceive a visit without bringing something (material support) as support. The findings have been very significant in the development of an intervention strategy to be implemented during the second phase of the pilot study

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/41 - The Distribution and Intensity of Parasite Sequestration in Comatose Malaŵian Children

KB Seydel ¹, DA Milner ¹, SB Kamiza ², ME Molyneux ³, TE Taylor ¹

Abstract

Objectives

Abstract

Background

With 14.4% HIV prevalence among Malawians aged 1549 years and limited antiretroviral (ARV) clinic facilities, interventions in ARV clinics are focusing more on physical than psychological care.

Objective

To determine the prevalence of psychological distress and associated factors among people living with AIDS (PLWAs) attending ARV clinics in Mzuzu City.

Methods

A cross-sectional survey was done among PLWAs attending ARV clinics at Mzuzu Central and St. John's hospitals. 440 clients were sampled using systematic random sampling. The “Self Reporting Questionnaire” (SRQ, social-demographics and other questions on predictive factors for psychological distress were administered verbally to participants. Ethical clearance was sought from COMREC. SPSS and STATA were used to analyze study data.

Findings

14.4% of this sample was distressed with 4.5 % having suicidal feelings. Young age, being female, low education, joblessness, poor social economical status were demographic characteristics associated with psychological distress. Shorter duration on ARVs and non-availability of psychological support were two of the clinical factors that were predictive of psychological distress. Logistic regression did not find any statistically significant interaction or confounding brought by age, sex or social-economic status.

Conclusions

There is unrecognized psychological distress among PLWAs in ARV clinics. Health care providers should consider the needs of the patient (both physical and psychological) in a comprehensive way.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/46 - The Relative Validity of an interactive 24 hour recall in estimating Energy And Nutrient Intakes in 16-month old Rural Malawian Children

CM Thakwalakwa ^1,², JC Phuka ^1,², M Manary ³, KM Maleta ¹, P Ashorn ^2,⁴

Abstract

Objectives

To determine the relative validity of a structured interactive 24-hour dietary recall in assessing energy and selected nutrient intakes for 16-months old rural Malawian children. To compare nutrient adequacy in the children as assessed by the two methods.

Setting

26 villages in the Lungwena Health Center's catchment area.

Materials And Methods

Cross-sectional study. Food consumption estimated on the same day using Weighed Food Record (WFR) and structured interactive 24 hour recall (24-hr) methods. SPSS version 10.1 calculated mean absolute intakes and percentage RDA for energy and selected nutrients from the two methods to determine agreement.

Results

169 16 months old children participating in a 12-month complementary feeding clinical trial called Lungwena Child Nutrition Intervention study (LCNI-3) participated.

The 24-hr gave 4% −53% higher mean absolute intakes than the WFR. Energy had the lowest (3%) followed by iron (17%) and fat (25%) while Calcium had the biggest difference (53%) followed by Vitamin C (37%).

Both methods showed that <25% of the children meet RDA for energy, iron, zinc, calcium and vitamin A. >50% of the children met at least half of the RDA for energy, protein and vitamin C as shown by both methods.

Conclusions And Recommendations

The 24-hr can replace the WFR to assess energy and iron intakes of infants in rural areas in developing countries. However it cannot replace the WFR to assess intakes of proteins, vitamin A, calcium, fats, vitamin C and zinc as it tends to overestimate. The 24-hr can replace the WFR in assessing nutrient adequacy for energy, protein, iron, calcium and vitamin A and vitamin C in rural children.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/47 - The Role of Il-17 in the Pathophysiology of Invasive Pneumococcal Disease

LA Mankhambo ¹, ED Carroll ³, P Balmer ², S Nkhoma ¹, G Jeffers ³, M Guiver ², CA Hart ⁴

Abstract

Background

Abstract

Objectives

The objective of this study was to improve understanding of transactional sex by examining it in the context of rural Malawi. Transactional sex is thought to be a major driver of the .AIDS epidemic in Malawi as elsewhere in the region. In the literature on AIDS, transactional sex is seen as a degraded form of sexual expression forced on vulnerable women by economic desperation. In this study, we ask how the exchange of sex for money and/or gifts is perceived by rural Malawians.

Study Methods

The method used to collect data was ethnography. A total of 21 Form Four graduates living in or around the three research sites of the University of Pennsylvania's Malawi Diffusion and Ideational Change Project (MDICP) kept notes on public conversations they overheard in the course of their daily lives. The MDICP now has approximately 600 field journals, covering the period from 1999 to the present. The journals were coded and analyzed using a qualitative analysis software.

Results

Our analyses show that the exchange of sex for money and/or gifts is widespread, but that the standard explanation that links transactional sex to prostitution and emphasizes the exploitation of poor, vulnerable women by wealthier, more powerful men, misses a great deal of what motivates and sustains such sexual patterns. We find that the understanding of transactional sex in rural Malawi can be better understood by seeing it as but one of the many ties of unequal exchange in which Malawians engage to provide security in an uncertain world. These ties are held together by a powerful moral ethic of redistribution and reciprocity, and are far more pervasive than transactional sex: they include, for example, the patron- client relationships through which the wealthier find jobs and other opportunities, and through which the poor exchange farm labor and food as well as other forms of support in times of emergency.

Conclusion

We believe that the pervasiveness of multiple sexual partnerships is better understood not as driven by either men's nature or by women's poverty. Both men and women are seeking to establish ties of dependence that provide a form of social insurance. These ties mitigate uncertain risk for both men and women by binding relatives, patrons and clients--at every social stratum, and in many of life's activities—in a web of ties held together by a moral ethic of redistribution and reciprocity.

Recommendations

If transactional sex is understood as a form of redistribution and reciprocity that has benefits for those seeking to mitigate social as well as economic insecurity, it suggests that simply labeling it as prostitution by needy women, ignoring the role of men, and advocating the end of this practice will not be sufficient. Researchers and policy makers need to have a more realistic understanding of the circumstances that make dependence on others an important form of social insurance even among the wealthy, and the major form of social insurance for the poor.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/51 - School, Sex, and Social Networks among Unmarried Women in Rural Malawi

M Poulin ¹

Abstract

Objectives

To compare HIV risk behavior of unmarried school-going girls with school-leavers, and to identify whether schooling is protective or not.

Methodology

This paper examines the links between schooling, sexual activity, and AIDS-related social networks among unmarried women (aged 15–24 years) in three rural districts of the country as part of the Malawi Diffusion and Ideational Change Project (MDICP). The present findings come from survey data, in-depth interviews, and HIV biomarker status from a sub-set (N=258) of the 2004 youth sample.

Results

The survey data show that school-going girls report consistently lower levels of sexual activity than do women who are no longer schooling, even after controlling for age, wealth, and region. Yet AIDS-related knowledge and attitudes do not differ across schooling status, suggesting that the difference between the sexual behavior of schoolgoers and non-goers is due not to the content of the school instruction. The qualitative data support the findings from the survey. Both show that a number of individual and social factors influence sexual choices, but these factors differ for girls who are in school versus girls who are not. Schoolgoers' networks are primarily composed of other girls who are also in school, and they encourage each other to avoid zibwenzis. In contrast, school-leavers' social networks are primarily composed of other, non-school-going girls, who seek marriage and are thereby more likely to have had sex.

Conclusions and Recommendations

Attending school is protective against the risk of HIV infection. This is not due to formal education itself, but because schoolgirls see a better future for themselves by delaying marriage and avoiding premarital sex. These findings have important policy implications, especially in its emphasis on how social networks modify using AIDS-associated information.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/52 - Effect of Counseling HIV Infected Mothers on Early Breastfeeding Cessation and Abrupt Weaning in Blantyre, Malawi

FG Zulu

Abstract

Introduction

Breastfeeding (BF) is a major risk factor for mother-tochild HIV transmission. However, breastfeeding is universal and replacement feed is considered not feasible because of cost and availability. For an HIV infected mother stopping breastfeeding altogether is the surest way of eliminating breastfeeding associated HIV infection. MOH/WHO is recommending informing HIV infected mothers to make informed decisions regarding breastfeeding. Guidelines and other tools have been developed for this purpose for resource poor settings

Aim

The aim of our study was to determine the effectiveness of intensive education and counseling of HIV infected women to stop breastfeeding early and abruptly.

Methods

This was part of a large clinical trial investigating the use of Post Exposure Prophylaxis for Infants to reduce HIV MTCT through breastfeeding. All HIV infected women who enrolled in the main study received counseling and education to stop breastfeeding at 6 months based on WHO/MOH guidelines. Women were continuously counseled to exclusively breastfeed and wean abruptly. They were instructed on food replacements, preparation and feeding practices. They were interviewed at each visit whether they actually decided to stop breastfeeding.

Data on various factors demographic, Socio economic and health status were collected through interviews using standard questionnaire to determine whether these factors were related with ability to stop breastfeeding early.

Results

1100 women had enrolled in the study, 800 of them had reached 6 months of follow up. Their mean age was 26.7 (18–44) years. More than 95% were married. Of the 800 mothers 81.1% weaned their babies abruptly at 6 months. Factors such as partner education, ability to read and history of involuntary weight loss were significantly associated with early breastfeeding cessation.

Conclusion

In this study, education and counseling of HIV infected women to stop breastfeeding early was effective. Implementation of early breastfeeding cessation programs must be considered.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/53 - Complementary Feeding with Fortified Spread Reduces the Incidence of Severe Stunting among 6–18 Month Old Rural Malawian Infants

J Phuka ¹, K Maleta ¹, C Thakwalakwa ¹, M Manary ¹, A Briend ¹, P Ashorn ¹

Abstract

Background

Childhood growth failure and incidence of malnutrition are very common in Malawi. Fortified spreads have proven effective in malnutrition rehabilitation and may also offer a means to prevention.

Objective

To compare growth and incidence of malnutrition among infants receiving long-term dietary complementation with either ready to use fortified spread (FS) or maize-soy flour (Likuni-Phala).

Methods

A randomised, controlled, single-blinded clinical trial. Sixmonth-old healthy infants were randomised to receive daily supplementation for 12 months of either 25g FS (FS25 group), 50g FS (FS 50), or 72g of LP (LP). All supplements were fortified with micronutrients. Daily dose for FS50 and LP contained 256 kcal energy, FS25 half of that. Outcome measures included changes in weight and length and incidence of severe malnutrition.

Results

182 infants started and 168 completed the intervention. The mean (SD) weight gain in the FS25, FS50 or LP groups were 2.36 (0.61), 2.45 (0.78), and 2.36 (0.60), respectively. Comparable mean (SD) gains in length were 13.0 (2.7), 13.0 (2.4), and 12.6 (1.7) cm. The incidence of severe underweight (WAZ<−3) was 18%, 19% and 15%, severe stunting (HAZ<3) 3%, 0% and 12%, and severe wasting (WHZ<-3) 5%, 2% and 2%, in the FS25, FS50, and LP groups, respectively. Compared to infants receiving LP, those getting FS had a 12% (95% CI 4 to 20%) lower incidence of severe stunting.

Conclusions

One-year-long complementary feeding with FS has at maximum modest effect on the infants' average growth, but it seems to markedly decrease the incidence of severe stunting.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/54 - Incidence and Serotype Distribution of Invasive Group B Streptococcal Infection in Infants in Blantyre, Malawi

KJ Gray ¹, SL Bennett ², N French ¹, AJ Phiri ¹, SM Graham ^1,²

Abstract

Objectives

To serologically characterize the Group B streptococci (GBS) collected from infants over a 14-month period in Queen Elizabeth Hospital, Blantyre, Malawi (QECH), to investigate the patient characteristics of disease from whom the GBS were isolated and to estimate the burden of invasive GBS in infants in Blantyre.

Materials And Methods

Over the period May 2004–June 2005, all Group B streptococci isolated from blood cultures and cerebrospinal fluids in Queen Elizabeth Central Hospital, Blantyre, Malawi were collected, serotyped and had antibiotic susceptibilities performed. Clinical data and outcome were collected retrospectively on infants with invasive disease. Rates of disease were estimated for Blantyre district.

Results

GBS were isolated from 57 infants. The overall incidence was 1.8 cases per 1000 live births; early-onset disease (0–6 days after birth) was 0.92 per 1000 live births; and late-onset disease (7–90 days after birth) was 0.89 per 1000 live births. Disease presented as sepsis (44%), meningitis (37%), pneumonia (5%) and was uncertain in 14%. There was a high overall case-fatality rate of 35%. Serotypes la and lll were responsible for 77% of disease with serotypes lb, ll and V making up the remainder. All isolates were susceptible to penicillin but 21% were resistant to erythromycin and 96% resistant to tetracycline.

Conclusions And Recommendations

This suggests that neonatal GBS disease in Malawi is similar to that in the developed world both in terms of the microbiology and disease presentation. The case-fatality is much higher suggesting that an intervention in this country could have a substantial impact on neonatal morbidity and mortality. Maternal vaccination in this setting merits further evaluation.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/55 - Implementation of “One StopSupermarket Approach” in Prevention Of Parents-to-child HIV Transmission in Thyolo District Hospital

J Kazima ¹, M Moens ¹, L Buhendwa ¹, M Massaquoi ¹

Abstract

Objectives and Scope

The PMTCT program at Thyolo District Hospital has been implemented since 2002 in the context of an extremely high HIV seroprevalence (23%) in antenatal clinic (ANC) setting. The objective of this new strategy is to increase access to PMTCT services for positive pregnant mothers at Thyolo District Hospital.

Methodology

The study was conducted looking at data from August 2005 – July 2006. We compared the uptake of pregnant mothers for PMTCT services before and after the implementation of the “one stop supermarket approach”.

All mothers received group motivation talks in the waiting area and are then offered HIV and syphilis testing and post test counseling, at the first contact.

All the services are integrated in the normal antenatal package. Opportunistic infection management, Cotrimoxazole prophylaxis and ART are offered to HIV+ mothers and the service is provided within the PMTCT infrastructure. The nutritional supplement (RUTF) is given to all HIV positive pregnant mothers and children during the post-natal period. This will help to improve the adherence to the program till 18 months.

Lessons learnt

With the current “One Stop Approach”, the uptake of PMTCT services is high (up to 98% of new ante natal mothers). Before the uptake was on average 70%.

There is an excellent “patient flow circuit” not allowing queues and therefore minimum waiting time at all levels. Accelerated HIV testing and counseling along with NVP administration is done in the maternity unit and limits missed HIV-positive deliveries occurring in the labor ward. The staff (NGO and MOH) who were complaining before of overwork are now highly motivated and demonstrating good team spirit.

Conclusion

“One stop shop service” in a hospital setting seems to be a positive way forward to increase uptake in PMTCT programmes. . The mothers receive the maximum package of activities on the same day of visit and at the same site and by the same nurse.

Recommendations

The pregnant mothers offered “routine counseling and HIV testing” should always be aware of their right to “OPT-OUT”. All the positive mothers attending the Hospital based PMTCT program should have the option of continuing their care and possible initiation of ART at health centers that maybe closer to their home. The implementation of “one stop market” strategy should be extended to health centers where PMTCT is already run.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/56 - Treatment Of Kaposi Sarcoma attending the Continuum of Care Clinic in Thyolo District Hospital

F Samura ¹, D Misinde ¹, C Foncha ¹, O Pasulani ¹, G Malata ¹, D Ras ², K Mbewa ², M Massaquoi ¹, R Zachariah ³

Abstract

Background

With HIV/AIDS epidemic KS incidence has arisen dramatically. KS generally regresses under ART and patients, who have a good prognosis, may benefit from ART alone.

Objectives

To document the characteristics and initial outcomes of HIV-patients on treatment for Kaposi sarcoma (KS), in a resource poor setting.

Method

Retrospective descriptive study using routine data. Inclusion - All patients with clinical diagnosis of KS receiving chemotherapy and/or HAART between 2003 until September 2007. Regular ART and chemotherapy outcomes were recorded.

Results

593 patients diagnosed and treated for KS. Mean age 33.7 years and male/female ratio 2/1. Ever stated ART 326/593 (54.9747%). <50 CD4 cells/mm³ 87, >50 <199 cells/mm³ 141, >200 < 499 cells/mm³ 76 and > 500 CD4 cells/ mm³ 8. The mean CD4-count 132.1 cells/mm³. ACTG KS classification: T0S0 4/593 (0.67%), T0S1 7/593 (1.18%), T1S0 44/593 (7.42%), T1S1 155/593 (26.14%) and 383/593 (64.59%) not staged. WHO clinical staging: stage 4 226 and stage 3 60. ART regimen: 1st line ART 300/326 (90.03%), alternate 1st line 26/326 (7.97%). Chemotherapy (bleomycin): Ever started bleomycin 192/593 (32.38%, completed 10 cycles 49/192 (25.52%), completed 20 cycles 12/192 (6.25%). Treatment outcomes: Alive 206/326 (63.19%), death 74/326 (22.70%), defaulters (4.91%), stopped treatment 3/326 (0.92%) and transferred out 13/326 (3.99%). Side effects: Cutaneous hypersensitivity 1 and PSP 22.

Conclusions And Recommendations

KS represents a significant burden of disease in patients with advanced AIDS and their outcomes are poor than overall patient outcomes. However, all cases followed up showed a remarkable improvement of the lesions, oedema, lymph nodes and general condition.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/57 - Conjugate Pneumococcal Vaccines as Secondary Prophylaxis in HIV-infected Adults

N French ¹, T Mdezo ¹, G Mwafulira ¹, H Longwe ¹, N Mtunthama ¹, R Malamba ¹, M Kunkeyani ¹, S Gordon ¹

Abstract

Objective

To determine the efficacy of a seven valent pneumococcal conjugate vaccine (CPV) in Malawian adults to prevent recurrent invasive pneumococcal disease (IPD).

Background

IPD is a leading cause of ill-health globally in HIV-infected adults. Measures to reduce pneumococcal disease in this group are required. Pneumococcal plain polysaccharide vaccines are ineffective in HIV-infected Africans but the newer CPV's are untested in adults. A double blind randomised placebo-controlled trial was started in Blantyre in 2003 to evaluate this vaccine.

Methods

Adults, who have suffered an IPD event (pneumococcus cultured from blood culture, CSF or other sterile site) and are identified on the adult medical wards of the QECH, are invited to participate. Following informed consent participants receive two doses of either seven-valent CPV or saline placebo. Participants are then followed up every 3 months or when unwell for evaluation. The primary endpoint of the study is vaccine serotype IPD. A secondary prophylaxis trial was chosen in order to minimise the size of the study - estimated sample size 420 person years of observation (c.f. primary prophylaxis study 3000 pyo).

Findings to date

As of end August the study has recruited 411 participants (87% HIV positive) and accumulated 409 pyo. There have been 56 IPD events in 42 individuals (first event rate 102/1000pyo, all HIV positive) of which 21 (38% of all IPD events) have been vaccine serotypes in 20 individuals. 102 (25% of total, rate 249/1000pyo) participants have died. The study has been recommended to continue by the data monitoring committee until October 2007 on the basis of lower than predicted numbers of primary end-points - a priori end point target was 42. Recruitment will continue until June 06.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/58 - The Bangwe Project - The Impact ofAntiretroviral Therapy

C Bowie ¹, N Gondwe ¹, MC Bowie ¹

Abstract

Background

Careful assessment and follow up of patients receiving home based care in a defined population of Bangwe, Malawi provides details of the frequency and severity of common symptoms and survival. This information can be used to assess the impact of antiretroviral therapy.

Methods

Mortality and the incidence, duration and severity of common symptoms of patients in a defined population receiving home based care were measured over a 42 month period. Two cohorts were constructed from the data, one starting in January 2005 when ART became available, and the other preceding the advent of ART and used to provide “historical controls”.

Results

886 patients, of whom 446 died, were studied. A quarter of patients died within two months of being first seen. About a third of the patients were unable to care for themselves on first assessment and two thirds had stage 4 disease. Half (58%) were malnourished with a Body Mass Index (BMI) < 18.5kg/m2. Most patients had a mixture of symptoms at presentation.

91 patients were on ART. There was no effect on the survival of the 2005–2006 cohort for whom ART was available compared to the 2003–2004 cohort.

Conclusions

The lack of impact of ART on survival could be due to either limited effectiveness of triple therapy in this group of patients with relatively advanced disease or due to those receiving ART having good survival in the early period after diagnosis even without ART. The impact of ART may take time to appear. This is of concern as first line ART has limited durability and Malawi may fail to benefit from the costly intervention.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/59 - Isolation, Identification and Speciation of Mycobacterial Organisms Associated with Tuberculosis In HIV-infected and Uninfected Adult Patients

IOO Komolafe ¹, SR Meshnick ², EE Zijlstra ³, RS Mkakosya ¹, A Van Rie ⁴, FM Salaniponi ⁵

Abstract

Objective

To isolate and characterize mycobacterial organisms associated with tuberculosis in cohorts of HIV-infected and uninfected adult patients at the QECH, Blantyre.

Methodology

Sputum specimens from index TB patients were subjected to mycobacteriology and culture. Blood samples from enrolled patients were also subjected to HIV serology using two spot tests and a tie-breaker as recommended by the National AIDS Commission.

Results

Sputum specimens from 136 patients were studied. 21 (15%) were acid-fast bacilli (AFB) positive but negative on culture and 18 (86%) patients in this group were HIV reactive. Of the 115 patients with smear and culture positive specimens, male (57) and female (58) were in equal proportions of 50% and 88 (76%) were HIV positive. Of the 155 mycobacterial isolates recovered from the participants, 94 (61%) grew on glycerol-based LJ slant, 57 (37%) on pyruvate-based slant while 40 (26%) of these numbers grew on both slants incubated at 36° –37°Celsius. Only 4 (2.5%) isolates suggestive of Mycobacterium xenopi grew on slants kept at 440C and all 4 patients were HIV positive. There is no significant difference between HIV positive patients which constitute 76% of the study population and produced 74% of the isolates and HIV negative patients (24%) producing 26% of the isolates. Serpentine cord formation, a cultural characteristic of Mycobacterium tuberculosis complex was detected in 66 (92%) of the 72 isolates examined so far.

Conclusion

This is an on-going study but preliminary results are pointing to M. tuberculosis complex as the most preponderant mycobacterial organisms associated with both HIV positive and negative TB patients.

Detailed organism speciation by a combination of biochemical and PCR techniques including antibiotic sensitivities of M. tuberculosis isolates are yet to be carried out.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/60- Where are Our Graduates?

EE Zijlstra ¹

Abstract

Objectives

To assess where the graduates of the College of Medicine are working and in what capacity.

Methods

Simple audit by obtaining information from graduates, classmates and College staff.

Results

Between 1992 and 2006, 232 students graduated (78% male) with a mean of 17 graduates per year. Of these 8 died and 18 are still completing their internship. The remaining 206 are working as follows: specialist [22 (11%)], post-graduate student [60 (29%)], PhD student [5 (2%)], medical officer [94 (46%)], administrator [14 (7%)]; not practising are 4 (2%) and for 7 it is unknown (3%). Currently in Malawi are 123 (60% of total); of these 48 work for the Government (MOH 45, NAC 3), 27 in the COM, 20 in mission hospitals and 11 in private practice. Currently abroad are 83 (40%); 55 are in post-graduate training. 40 (19% of total) are in the UK, 22 (11%) in the RSA and 8 (4%) in the USA; 24 are unlikely to return to Malawi.

Conclusion

The majority of COM graduates are in Malawi and practising. Of those abroad most are in postgraduate training and expected to return.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/61 - Acceptance of Door-to-door Voluntary Counseling and Testing in Rural Malawi

S Yeatman ¹, P Fleming ²

Abstract

Objectives

To evaluate the acceptability of door-to-door voluntary counseling and testing (VCT) in rural Malawi and discuss its relevance for HIV research and HIV testing policy.

Methods

During June and July 2007, as part of the data collection for the fourth wave of the Malawi Diffusion and Ideational Change Project (MDICP), we conducted door-to-door voluntary counseling and testing of our random population-based sample of 3000 men, women and adolescents in the Balaka, Mchinji and Rumphi Districts of Malawi.

Results

Preliminary findings from two sites suggest that approximately 90% of persons in the sample approached for VCT agreed to be tested with rapid tests in their homes. Of those tested, approximately 98% chose to receive their test results immediately. Qualitative interviews conducted with a sub-sample of those who accepted and refused an HIV test reveal that most individuals were comfortable with having an HIV test at their home. Those who refused an HIV test generally reported that they did not want an HIV test because they felt that they were not at risk for HIV or that they already had been tested for HIV in the near past.

Conclusions and Recommendations

Research projects that intend to measure HIV prevalence of populations or samples can learn from the high acceptance rates and demand for door-to-door rapid testing observed in this rural Malawian cohort. Additionally, while door-to-door VCT is not logistically feasible on a large-scale in resource-poor settings, the success of home-based rapid testing suggests that people want to be tested and that alternative provision of counseling and testing may increase its use. Mobile clinics have been popular in other rural sub-Saharan settings and may provide an environment that lessens logistical barriers and perceived confidentiality threats commonly associated with testing at clinics. This conclusion is consistent with the success of the July 2006 National Testing Week in Malawi and suggests that a move towards less rights-based policies such as routine testing in health centers may be premature.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/62 - ABO Blood Group Phenotypes and Placental Malaria

EL Senga ¹, MP Loscertales ², BJ Brabin ^2,^3,⁴, PN Kazembe ⁵, KEB Makwakwa ¹, GN Liomba ¹, C Dzamalala ¹

Abstract

Objective

To assess the association between the ABO blood group (bgp) phenotypes and the placental malaria (plmal) infection.

To investigate if pregnancy outcomes are associated with ABO bgps.

Materials And Methods

The study was conducted between Feb – June 2004 and Jan – July 2005 at Montfort Hospital in Chikwawa - Malawi. A total of 647 women were enrolled. Maternal peripheral and cord blood samples were obtained after delivery. Using standard haematological methods the ABO bgps were typed with commercial anti-serum reagents. Placenta tissues (1cm cubes) were obtained and fixed in 10% formalin, processed and were examined microscopically by three observers. Placental histology was classified according to Bulmer's classification.of acute, chronic, past and no infection.

Results

The ABO phenotype distribution was 146 (A), 21(AB), 161(B) and 309(O). Of the 647 women, 206 (31.8%) were primigravid (PG) and 441(68.2%) were multigravid (MG). Overall, 235(36.4%) had plmal infection and 401(60%) had no infection and 11 had no data. 124(19.2%) had active infection and 111(17.2%) had past infection. The risk of plmal infection was higher in PG: OR 2.15, 95% CI 1.53- 3.01, P<0.001. Blood group O was significantly associated to a higher risk of acute plmal in primiparae (PP) [OR 2.18, 95% CI 1.05–4.55, Chi-sq 5.17, p=0.02] and to a lower risk in the multiparae (MP) [OR 0.59, 95% CI 0.35– 1.00, Chisq 4.29, p= 0.04]. Babies of O group mothers in their first pregnancy have an increased risk of falciparum placental malaria. Mean birth weight was higher in mothers with than without blood group O, (p=.05) , (O: 2808gms PP vs 3098gms MP, p<0.001; non-O: 2828gms PP vs 3017gms MP, p<0.001). Mean newborn ponderal index was higher in O bgp MP (p=0.007).

Conclusion And Recommendation

Blood group O is a risk factor for increased malaria risk in first pregnancies and reduced risk in later pregnancies in these Malawian women.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/63 - Cryptosporidium in Paediatric Patients in Chikwawa and Blanytre Districts, Malawi

T Morse ^1,³, RAB Nichols ², AM Grimason ³, H V Smith ²

Abstract

Background

Cryptosporidiosis can be a severe, protracted and debilitating disease in the immunocompromised, and may lead to death due to the lack of current effective treatment. Risk factors and the epidemiology of disease must be understood to reduce exposure to infectious Cryptosporidium oocysts. This requires species identification as well as assessment of potential risk factors that lead to infection.

Objectives

To identify species of Cryptosporidium infecting children in both urban and rural Malawi, and to assess the potential sources of infection with a view to identifying suitable controls to reduce infection.

Methods

A hospital based study enrolled patients were enrolled from Queen Elizabeth Central Hospital under five OPD, Chikwawa District Hospital OPD/paediatrics/NRU and three remote health clinics in Chikwawa District. Stool samples procured from patients under the age of five years suffering from diarrhoea were examined using brightfield and epifluorescence microscopic methods in Malawi, with all putative positive samples subject to PCR-RFLP for species identification at the Scottish Parasite Diagnostic Laboratory (UK).

Positive cases in Chikwawa District were further included in a case control study to assess potential sources of infection. Data collected from each case and control included questionnaires, water samples, handwashing samples, animal stools, latrine surface swabs and stool samples from household members assessing 61 risk factors in total.

Results

Cryptosporidium spp. oocysts were seen in 5.9% (50 / 848) of samples, of which 43 amplified by PCR-RFLP indicating the following species: C. hominis, C. parvum, C. hominis / C parvum, C. meleagridis and C. andersoni. Seven samples could not be amplified by PCR. Seasonal variations were seen in disease prevalence, and age was a determining factor in contraction of infection. Wider species diversity was found in the rural setting. Risks were identified for all reported routes of Cryptosporidium transmission, namely person-to-person, waterborne, zoonotic and foodborne. The potential for faecal-oral transmission, zoonotic and water contamination were of particular interest.

Conclusions

Due to the significant health implications for the malnourished and immunocompromised from cryptosporidiosis, it is essential to develop, conduct and monitor suitable IEC materials to ensure community hygiene. This must include the participation of HSA's and VHC's if messages are be effective and practices are to improve.

Malawi Med J. 2007 Mar;19(1):39–65.

DCON/06/64 - HIV Point-prevalence amongst Malnourished Children admitted to Nutritional Rehabilitation Units in Malawi: Geographical & Seasonal Variations

SA Thurstans ¹, M Kerac ², K Maleta ², TW Banda ³, A Nesbitt ²

Abstract

Background

HIV/AIDS and malnutrition are inextricably linked and together represent a serious humanitarian and public health challenge in Southern Africa. However, their treatment has remained largely separate, and seasonal and geographic variations in presentation are often ignored.

Objective

The aim of this study was to collect information to guide integrated, targeted care.

Objective

To explore the perceptions and attitudes of PLWA towards HIV partner notification

Study methods

This was a cross-sectional study that utilized a semi-structured questionnaire administered to HIV infected persons living in Bangwe, a township in Blantyre, and participating in a home based care program run by the College of Medicine.

Results

A random sample of 75 HIV positive people was interviewed. More females (49) than males were represented in the sample. The mean age was 35.7 years (SD 9.6). Overall, 59% of the respondents perceived partner notification as necessary. 95% of participants indicated that they would be willing to participate in a partner notification program as long as it was confidential and voluntary. Provider-referral was the preferred method of partner notification. The major barrier to patient-led notification was fear of discrimination and stigma. Other barriers were fear of losing respect, shyness, fear of being abandoned, divorce and termination of a relationship.

Conclusion

This study has finds that partner notification is considered by many PLWHA an appropriate strategy to assist HIV prevention and control.

PERMALINK

Report on the 10th Annual College of Medicine Research Dissemination Meeting

Dr SM Graham

Roles

DCON/06/01 - National Survey of Surgical Activities At The District Hospital In Malawi

C Lavy

N Mkandawire

C Steilechner

A Tindal

S Chimangeni

Abstract

Objectives

Materials And Methods

Results

Conclusions And Recommendations

DCON/06/02-Accelerated HIV Testing for PMTCT in Maternity and Labour Wards is vital to capture Mothers at a Critical Point in the programme at district level in Malawi

JJ Beltman

M Fitzgerald

L Buhendwa

M Moens

R Zachariah

J Kazima

N Alide

Abstract

Introduction

Methods

Results

Conclusions

DCON/06/03 - Aetiology of Pulmonary Infections in Adult Patients admitted to the High Dependency Unit at Queen Elizabeth Central Hospital

TH Hartung

D Chimbayo

AJ Phiri

ME Molyneux

EE Zijlstra

Abstract

Objectives

Methods

Results

Conclusion And Recommendation

DCON/06/04 - Affordable Miniaturized Flowcytometric CD4+ T-Cell Enumeration, The Blantyrecount, For Clinical Use In Malawi

CA MacLennan

MKP Liu

S White

F Simukonda

J Bwanali

JJG van Oosterhout

MJ Moore

EE Zijlstra

MT Drayson

ME Molyneux

Abstract

Objectives

Setting

Materials And Methods

Results

Conclusion And Recommendations

DCON/06/05 - Anthropological And Bioethics Study of Clinical Research in Malawi

JM Mfutso-Bengo

M Mkunthi

V Jumbe

E Robson

F Masiye

M Molyneux

Abstract

Introduction

Objectives

Setting

Methodology

Results

Conclusions

DCON/06/06 - Anthropometric Characteristics of Underfive Children and Adults in Lungwena, Mangochi

DM Chilima

KM Maleta

Abstract

Objective

Setting

Material and Methods

Results

Conclusion And Recommendations

DCON/06/07 - Antibiotics for Prevention of Preterm Labour (Apple) Trial: A Randomised Controlled Trial