Morbidity and mortality from bacterial meningitis remains high. (13, 5, 7, 4) While delayed presentation, partial treatment, altered sensorium and pneumococcal meningitis are known risk factors for mortality, (2, 17) those for morbidity are not as well defined. (16) Neurological sequelae constitute some of the more common complications of childhood bacterial meningitis. (3, 8) They tend to occur when there is some delay in diagnosis and treatment and such situations are not unusual in developing countries like ours. Common neurological sequelae that have been described in childhood meningitis include problematic neuromotor development, learning disorders, ocular and visual problems, hearing loss, language and speech disorders, other cranial nerve deficits, seizure disorders, hydrocephalus and behavioural problems. (4, 3, 8, 6) Some of these sequelae occur early in the illness and are transient, whereas, others remain permanent and presumably contribute to the growing pool of children with disability in our environment.
Although subtle neurological deficits are generally found to be more prevalent than moderately severe or severe ones, Bedford et al (4) found a 10-fold increased risk of moderate or severe disability at the age of 5 years in English and Welsh children who suffered from meningitis in infancy when compared with children who did not. Such a growing load of children with disability will undoubtedly overstretch the limited resources available for childhood healthcare needs especially in resource-poor countries such as Nigeria.
The focus of meningitis studies in Africa (11, 12, 14, 15) has been on epidemiology. Other outcomes of the disease have largely been neglected. There is therefore a need to study the pattern of short and long - term neurological sequelae of children diagnosed with pyogenic meningitis. These sequelae also need to be related to presenting symptomatology and other prevailing factors at the time of diagnosis. This we believe would augment the dearth of information on the sequelae of this dreaded disease in our environment, serve as an impetus for prospective studies, as well as help in healthcare policy formulation.
This study is a retrospective hospital based review done at the Lagos University Teaching Hospital. The case notes of all children who were discharged following admission for meningitis between January 1 1991 and December 31 2000 were retrieved. Those with the final diagnosis of “pyogenic”, “bacterial”, or “purulent” meningitis were considered as cases while those diagnosed as tuberculous or viral meningitis were excluded from the review. The epidemiologic and clinical data of the cases were recorded in a data entry sheet drawn up by the researchers. Significant information collected include biodata, presenting symptoms and signs, duration of symptoms prior to presentation in our hospital as well as the treatment given, cerebrospinal fluid analysis results, blood culture and serology results, serum electrolyte and urea results, antibiotic and other therapeutic agents used, as well as the duration of hospitalization. In addition, all neurological complications recorded in the case notes at the time of discharge were entered into the data entry sheets.
All data was recorded and analysed using the SPSS statistical package. Associations were tested by the chi test, while students t test was applied to continuous variables. Appropriate non-parametric tests were used where applicable. Significance was set at p<0.05.
Forty - nine 49 children diagnosed with acute pyogenic meningitis survived their illness. Their ages ranged between 1 day and 120 months with a median of 6 months at presentation. Majority (23) of the children were aged between 1 and 11.9 months, 11 were neonates, 9 were aged between 12 and 59.9 months while nine were 60 months and older. Neurological sequelae were documented in 32 (65.3%) at discharge. Table I shows a breakdown of the various problems identified. They include hypertonia, hydrocephalus, speech and language disorders, hearing problems alone, seizures, mental retardation, visual defects and behavioural problems such as hyperactivity. Of the 10 children with hypertonia, 2 each had spastic quadriplegia and hemiplegia respectively. Three (9.9%) of the 6 children with visual defects were diagnosed with cortical blindness while the remaining had deficits of cranial nerves associated with the eye. In 17/32 (53.1%) multiple deficits were documented. Neuro-motor abnormalities topped the list of sequelae whereas when combined, auditory and speech problems predominated. Generally neonates tended to have fewer sequelae but this was not statistically significant. Beyond the neonatal age group, sequelae were commoner in infancy with a two or more times tendency to occur than in other age groups, but showed no statistical significance - Figure 1.
There were no significant differences in the various risk factors assessed when children with sequelae were compared with those without. We found that children with sequelae were older at presentation, had a longer duration of treatment before arrival at the LUTH, and stayed longer in hospital although the difference was not statistically significant - Table II. There was a tendency for more of the children with sequelae to present with fever, to have had more episodes of convulsion and features of meningeal irritation - Tables III and IV. Likewise children with neurological sequelae tended to have a higher total peripheral white blood cell count, to have had a higher cerebrospinal fluid protein level, and to have required more anticonvulsant and steroid therapy - Tables V and VI.
Neurological sequelae though undesirable remain common consequences of pyogenic meningitis in childhood, a fact further borne out by this study in which 32 of 49 (65.3%) children survivors of meningitis exhibited obvious neurological sequelae at the time of discharge from hospital. The prevalence of sequelae in our group of children is a lot higher than figures reported from various studies including those from other developing countries. It more than doubles prevalence figures from Norway (12) (15.2%), Edo State, Nigeria (1) (15.5%), Jordan (7) (20%) and Ghana (6) (22%). It is also much higher than figures from India (5) (39%) and South Africa (10) (46%) thus suggesting that there may be certain peculiar risk factors in our environment. We suspect that such factors may include rampant self-medication, uncontrolled access to antibiotic use by the general populace, and poverty, all resulting in prior partial treatment and a delay in presentation for adequate care.
Several risk factors are known to influence the outcome of meningitis, chief amongst which is the aetiologic agent. An assessment of some risk factors excluding bacterial causative agents in this study showed no significant relationship between any of them and the occurrence of neurological sequelae. There was a general tendency for children with sequelae to be older on admission : have received prior treatment before arrival; require antibiotic treatment for a longer; exhibit more neurological symptoms and signs at the time of presentation; have a higher total white blood cell count; have a higher CSF protein level and a higher CSF blood sugar ratio; and require use of more anticonvulsant /steroid therapy. However, none of these factors showed a significant relationship with the development of neurological sequelae. The absence of any significant relationship between these factors and the development of neurological sequelae presumably may be a function of the small numbers of cases in the study. Assessing the evolution of these deficits over time to determine the burden of residual deficits would have been useful but these children were lost to follow up as soon as they were discharged from hospital. This is a common finding in our environment and is a function of traditional cultural and religious beliefs where most disabilities are supposedly caused by supernatural means and as such cannot be treated by orthodox medicine.
As opposed to other studies where hearing deficits constitute the most common neurological sequelae of meningitis, disorders of tone and motor topped the list in our study. This seemingly less number of children with hearing deficits can be explained by the fact that only children who were obviously hard of hearing were identified. It is very likely that those who required special tests to diagnose their hearing loss were missed, as facilities to detect hearing loss especially in infancy did not exist in our institution then. In general however, the neurological sequelae exhibited by the survivors of meningitis in this study are similar to those described in young children admitted with bacterial meningitis in other parts of the world. Similarly, they tended to be commoner in infancy and occurred in multiples.
In conclusion, bacterial meningitis constitutes a ready source of physical and sensory disabilities in childhood and every effort must be made to ensure its prevention. While primary prevention of meningitis by immunization against the three most common causative agents of the disease viz. streptococcus pneumoniae, haemophilus influenzae and niesseria meningitides is ideal, there is need also to emphasize secondary prevention of disabilities by prompt, adequate and appropriate treatment. In addition the early recognition of disabilities is important to plan adequate rehabilitative services.
Table I: Pattern of neurological sequelae among survivors of meningitis
|Identified deficit||No (%)|
|Speech and language disorders||8 (25.0%)|
|Hearing problems alone||8 (25.0%)|
|Mental retardation||7 (21.9%)|
|Visual defects||6 (18.8%)|
|Behavioural problems||1 (3.1%)|
Table II: Clinical characteristics of survivors of meningitis
|Symptoms||Neurological sequelae present (N=32)||Neurological sequelae absent (N=17)||p value*|
|Age in months (SD)||18.0 (27.6)||16.4 (28.9)||0.5|
|Duration of treatment before hospitalization in days (SD)||12.7 (13.8)||4.7 (3.7)||0.3|
|Duration of hospitalization in days (SD)||24.5 (13.7)||22.5 (15.5)||0.7|
* Mann-Whitney U and Kruskall Wallis tests
Table III: Presenting symptoms among survivors with meningitis
|Symptoms||Neurological sequelae present (N=32)||Neurological sequelae absent (N=17)||p value|
|Fever||29 (93.5%)||17 (94.4%)||1.0|
|Convulsion||22 (71.0%)||11 (61.1%)||0.5|
|Irritability||9 (29.0%)||8 (44.4%)||0.2|
|Headache||3 (9.7%)||0 (0.0%)||0.3|
|Rash||1 (3.2%)||1 (5.6%)||0.6|
|Photophobia||1 (3.2%)||0 (0.0%)||0.6|
|*Others||9 (29.0%)||6 (33.3%)||0.7|
* others (coma=5, vomiting=3, anorexia=2, apnoeic attacks=2, diarrhoea=1, cough=1, squint=1, neck retraction=1)
Table IV: Presenting signs among survivors with meningitis
|Signs||Neurological sequelae present (N=32)||Neurological sequelae absent (N=17)||P value|
|*Seizures||21 (67.7%)||8 (44.4%)||0.2|
|Neck stiffness||10 (32.3%)||4 (22.2%)||0.5|
|Bulging anterior fontanelle||11 (35.5%)||3 (16.7%)||0.2|
|Kerning’s sign||6 (19.4%)||2 (11.1%)||0.7|
|Petechial rash||2 (6.5%)||0 (0.0%)||0.5|
* (generalised=13, partial=2, other=1, not stated=13)
Table V: Laboratory parameters of children surviving meningitis
|Symptoms||Neurological sequelae present (N=18)||Neurological sequelae absent (N=14)||p value*|
|Total White cell count (SD)||14766 (9463)||12246 (5957)||0.4|
|Lymphocyte %||33 (18)||35 (27)||0.7|
|Neutrophil %||64 (18)||63 (27)||0.9|
|CSF cell count||230 (277)||262 (456)||0.8|
|CSF protein||182.5 (162.9)||101.7 (100.6)||0.1|
|CSF:Blood sugar ratio||47.8 (26.1)||45.5 (22.3)||0.8|
* Mann-Whitney U and Kruskall Wallis tests
Table VI: Treatment given to survivors with meningitis
|Treatment||Neurological sequelae present||Neurological sequelae absent||P value|
|Penicillin + chloramphenicol or genticin||13||8||0.55|
|Cephalosporin ± genticin||8||9|
|Phenobarbitone ± any other drug||18||11||0.27|
|Hydrocortisone or dexamethasone||4||2|
| The authors retrospectively reviewed 49 children with acute bacterial
meningitis over a ten year period. A high proportion of the children had significant neurological sequelae ranging from weakness, sensory disturbance, hearing disorders, mental changes, recurrent seizures and
hydrocephalus. They correlated the prognosis to a number of variables. The only significant correlation was that of prior hospitalization and
treatment. They did not find a correlation between socio-economic factors or any other aspects.|
Prof AHMED I BHIGJEE
Subdepartment of Neurology
The Nelson R Mandela School of Medicine