The Growing Infant Meningitis Crisis in Ethiopia


MRI of infant with meningits
Infant with meningitis seen on MRI to have hydrocephalus and meningeal enhancement

Ethiopia is facing a growing crisis in infant meningitis compounded by a false reliance on the power of antibiotics to treat the problem without definitive diagnosis and follow up.

A growing problem
As you may know north eastern Africa including Ethiopia is part of the so-called “meningitis belt” with the highest rates of bacterial meningitis in the world. Just a couple a months ago I was attending a meeting with Ministry of Health with the new five pediatric neurosurgery centers of excellence in Ethiopia where a new growing epidemic of chronic meningitis seen throughout the country was discussed. Over the past year we are seeing more and more cases of infants on an almost daily basis presenting with progressive hydrocephalus who have diagnostic cerebrospinal fluid consistent with chronic meningitis. Our average census is 5 to 8 in hospital all the time whereas a few years ago it was only one or two. Many of these children have hospitalizations of 12 weeks or more. Many have died and many are left with significant cerebral disability which could have been prevented by earlier intervention and appropriate treatment lengths.

Systemic deficiencies
While this epidemic seems to be worsening, Dr. Abreha (Head of Pediatrics and Pediatric Neurologist at Mekelle University) and myself are concerned that deficiencies in facilities and procedures within the medical system may be contributing to it. Unfortunately throughout Ethiopia almost no lumbar punctures are now done to diagnose meningitis. A sample inquiry of pediatric residents and interns found that 75% of them had never done a lumbar puncture. Additionally many clinics and hospitals do not even have lumbar puncture trays. A false sense of security exists that powerful antibiotics given a short period time can will solve the problem

Secondly, the lack of ability to process or even receive an CSF specimen for analysis beyond 5 PM forces the treating physicians to start treatment without first obtaining a culture which is against world wide standard of culture first then start treatment. This seems very contrarian to the frequent theme of the pharmacy and laboratory professionals in Ethiopia that there is abuse of antibiotics.

A key analysis is determining the glucose, protein, and cell count in cerebrospinal fluid. Dr. Abreha recently gave a sample of tap water to the laboratory to be tested on the automated machine and it diagnosed multiple WBC consistent with meningitis. The equipment supplied to Ethiopian hospital is not capable of accurate diagnosis. Technicians can do manual counts but these are time consuming and labor intensive. This inaccuracy has greatly impaired our ability to determine if patients are infection free which they must be before they can undergo definitive treatment for hydrocephalus which is a ventriculoperitoneal shunt.

Thirdly, around the world there is much controversy about for how long antibiotic treatment should be given ranging from 10 days to 21 days. The problem is that partially treated meningitis patients who have not been cleared from the infection often do not have fever, meningeal signs, or other clinical findings except hydrocephalus. Many infants are briefly seen for nonspecific fever and receive short courses of antibiotics in Ethiopia without specific diagnosis being made and without adequate follow up. Many of these children we believe are harboring these low grade chronic infections leading to their late appearance at Ayder  Comprehensive Specialized Hospital. This creates a great dilemma as these children often require treatment of intravenous powerful expensive antibiotics from 21 days to in excess of six weeks or more until the infection is cleared by the demonstration of two negative cultures off antibiotics and normal cell counts. In addition in order for the shunt to work the protein has to be less than 150 mg/dl. Failure to diagnose an active residual infection before a shunt placement will only aggravate the infection leading to shunt removal and complications.

The treasure and future of Ethiopia is in her children. Meningitis appropriately treated early and followed can have a very low morbidity and disability outcome. The current situation if not acted upon will result in increasing medical costs, increasing disability, and increasing infant death which could be prevented with simple directed community and institutional action. 

Recommended Course of Action
1. Immediate upgrade of laboratory ability in hospitals to receive CSF specimens 24 hours a day for culture, gram stain, sensitivity.
2. Training of interns and residents in performing lumbar puncture and making sets available
3. Institute an Ethiopian wide policy of obtaining CSF before starting antibiotics. Children with seizures, lethargy, focal deficit, or signs of increased intracranial pressure would be emergently sent to a referral hospital for CT Scan or if possible if they have an  open fontanelle undergo an head ultrasound locally to rule out mass before lumbar puncture.
4. Make sure all children treated for meningitis undergo follow-up and that a repeat lumbar puncture is done after treatment. This may seem over kill but given the crisis happening it is the only way to prevent these chronic cases.
5. Funding of community service, public education, and research projects on this vital issue in cooperation with the regional health bureaus.
6. Training of health officers, nurses, and general practitioners in proper diagnosis, evaluation, treatment, and referral of children with meningitis and hydrocephalus.

Should Surgeons other than Neurosurgeons Treat Hydrocephalus in Africa?

An illustration of a ventricular peritoneal shunt

Traditionally surgery for hydrocephalus has been performed by neurosurgeons who undergo vigorous neuroscience basic training and supervised surgical experience of at least 5 years in length (longer in developed countries) after medical school. Recently there has been some discussion as to whether in African and other undeveloped countries general surgeons should be taught to perform ventriculoperitoneal shunts to treat hydrocephalus. The treatment of hydrocephalus should remain under the direction and in the hands of neurosurgeons.

Hydrocephalus as a medical condition has been recognized since the time of the ancient Greeks. The concept of surgery to treat hydrocephalus by diverting the flow of cerebrospinal fluid began in  1949, when Nulsen and Spitz implanted a shunt successfully into the caval vein with a ball valve. Between 1955 and 1960, four independent groups invented distal slit, proximal slit, and diaphragm valves almost simultaneously.

An estimated 750,000 people have hydrocephalus, and 160,000 ventricular peritoneal shunts are implanted each year worldwide almost always by neurosurgeons. About 56,600 children and adolescents younger than age 18 years have a shunt in place.

The incidence of hydrocephalus in Africa is estimated to be 145 per 100,000 which is three times higher than in the developed world. Thousands of these children will need surgical intervention, either ventricular peritoneal shunt or the newer, but still not clearly accepted as superior, endoscopic procedures. 

A survey conducted among African neurosurgeons in 1998 showed that there were 500 neurosurgeons in Africa; that is, one neurosurgeon for 1,350,000 inhabitants, and 70,000 km2. That number is significantly increased now but the exact current number is unknown. Worldwide the average is 1 neurosurgeon per 230,000 but in Africa it can be as low 1 per 9 million people. It is believed there are 700 neurosurgeons currently or about 1 per 1,238,000  people which is an improvement but still not nearly enough. Ethiopia currently has about 30 practicing neurosurgeons and will soon be graduating about 30 newly trained neurosurgeons per year. This means Ethiopia will need about 450 functioning neurosurgeons taking into account expected population growth. 

Although the training of neurosurgeons in performing ventriculoperitoneal shunts has become somewhat standardized via the World Federation of Neurosurgical Societies as well as international neurosurgical groups, the training of general surgeons to do this procedure has not been rigorously studied. There are very few publications about the results of general surgeons performing ventriculoperitoneal shunts but an a study from Kenya done in 2010 showed a significantly high complication rate of 65% with an infection rate of 9.1% and shunt malfunction rate of 11.1%. This was much different than reported by Dr. Warf , an American trained neurosurgeon who created a specialized center in Uganda, with a malfunction rate of 4%.

More recently endoscopic procedure to open the third ventricle to the cistern and coagulate the choroid plexus are gaining ground but not totally proven yet. These procedures clearly require specialized training and knowledge of anatomy of the caliber of a neurosurgeon and not a general surgeon.

The real issue was not really the shortage of surgeons but the bottleneck was lack of hospitals, operating rooms, and clinics. Additionally transportation to get healthcare is a real issue. Now Ethiopia has seen the light and has  three training programs for neurosurgery in Ethiopia and are graduating about 30 per year. Again the problem is we have more surgeons than facilities to operate in.

African governments will see the idea of adding shunt placement to general surgery as an easy fix. In reality they are already overworked. I have taught medical students and general surgery residents. Many people think ventriculoperitoneal shunts are the easiest procedure but I always tell my neurosurgery residents and fellows it is not. Decisions about when to shunt, is the shunt working, is it infected? require experience and training. Academic following of outcomes, techniques, epidemiology requires an academic neurosurgery program take the lead.

Unfortunately there is no shortcut to capacity building. We are now training neurosurgeons for other African countries as well as Ethiopia. Eventually we will need more than 450 which will take time.

Finally I would say that our approach to hydrocephalus is changing rapidly. For us and the Uganda group we are consistently reducing the number of shunts we are doing each year. Whereas in the past we did nearly two hundred it is now going to be less than 100 even though we cover 20 million plus population with the highest myelomeningocoel rate in the world. We now recognize that many neural tube defect newborns have low grade infections which require antibiotics sometimes over 21 days until the csf is clear. Many times their hydrocephalus stabilizes after a few fontanelle taps. Although the post infectious group is rising ( we are in the infamous meningitis belt of Africa) similar we have avoided shunting by similar close follow-up. Our shunt infection rate is currently 3% because we identify these chronic low grade infections. I shutter to think what would happen if general surgeons with little experience are let loose upon this situation.