CHAPTER 11
Stroke in the African context
Albertino Damasceno
Universidade Eduardo Mondlane, Maputo, Mozambique
11.0 Introduction
In this chapter we describe the results of an epidemiologic study to evaluate the hospital incidence rate of stroke, its main clinical characteristics, and the in-hospital and 28-day mortality. A standard methodology was used to allow comparisons with similar studies from other parts of the world, as well as with future studies in Maputo. Based in the capital of Mozambique, this prospective study involved hospitalized or ambulatory patients, all of whom were followed for 28 days; the vast majority of stroke diagnoses (including stroke type) were based on CT scan or necropsy. These new data highlighted a number of novel (to Africa) aspects around the epidemiology and impact of stroke.
11.1 An epidemiological study of stroke hospitalizations in Mozambique
Damasceno A, Gomes J, Azevedo A, Carrilho C, Lobo V, Lopes H, Madede T, Pravinrai P, Silva-Matos C, Jalla S, Stewart S, Lunet N. An epidemiological study of stroke hospitalisations in Maputo, Mozambique. Stroke 2010; 41:2463–69.
11.1.1 Study background
There are approximately 15 million new acute stroke cases every 12 months worldwide, with the majority of cases residing in low- and middle-income countries [58]. This is due to extended life expectancy and rising exposure to major risk factors of stroke, such as smoking, obesity, diabetes, and hypertension [59]. Stroke, which frequently reduces patients’ quality of life through cognitive and motor impairment and is ranked globally as the second leading cause of mortality, has recently become a concern in sub-Saharan Africa [59,60]. The impact of stroke in African countries is increasing due to a rising prevalence of hypertension associated with a very poor level of awareness and control. At the same time, there is limited access to health services and pharmacological treatment, poor standards of care, and delay in seeking acute hospital treatment [61].
Approximately one-third of the Mozambican adult population is hypertensive; of these, less than 10% receive pharmacological treatment [62]. With stroke risk factors adding a concern, 4.8% of the Mozambique population is obese [63] and approximately 30% smoke cigarettes [64]. In 1994, cerebrovascular disease was reported to be the sixth cause of mortality among individuals aged 15 to 59 years and was ranked first among individuals aged ≥60 years in Maputo [60]. In order to meet the increasing health service needs of the Maputo population, stroke surveillance is vital to plan future resource necessity [61].
11.1.2 Study aims
By utilizing the WHO’s STEPwise approach to Stroke Surveillance methodology [64], the study aimed to evaluate and estimate the new cases of stroke hospitalization, the epidemiological stroke event features, and the 28-day case-fatality rate and disability status in Maputo, Mozambique.
11.1.3 Study methods
11.1.3.1 Patient enrollment
During a full year, in the period of August 2005 to July 2006, all private or governmental hospital patients who had encountered a first-ever or recurrent stroke event and who had been residing in Maputo for ≥12 months were registered for the study. When new symptoms occurred in the same arterial distribution ≥29 days previously, or in a different arterial distribution caused by a former event <28 days previously, these were recorded as a recurrent stroke event [65]. Alternatively, events that occurred before the study period were referred to as old stroke events and were not included in the study analysis. When stroke-related events likely to be associated with HIV infection were suspected, the patients’ status of HIV infection was evaluated. However, all stroke events were recorded regardless of their HIV status.
11.1.3.2 Data collection
When stroke patients approached a private or governmental hospital, they were administered the STEPS Stroke survey, which included demographic and clinical questions. Trained interviewers were responsible for data collection and were positioned at the Maputo Central Hospital’s emergency department, as well as trained nurses in other emergency departments and trained physicians with cerebrovascular disease experience in medical wards.
First-ever and recurrent stroke events were self-reported by the patients as well as by the analysis of medical records, necropsy reports, and CT-scan images by physicians. Pharmacological treatment and assessment, biomedical tests, and in-hospital complications were recorded. Diabetes, hypertension, dyslipidemia, and AF were either self-reported or assessed through different analysis methods, including the prescription of pharmacological treatment, evaluation of questionnaires, and performing 12-lead ECG. Ethics approval was obtained, along with informed consent from all study patients.
11.1.4 Study findings
Figure 11.1 shows the schema of study participation. A total of 825 patients with reported stroke-related events were analyzed. Of these, 174 (21.1%) patients were excluded from the study; this was due to 40 (23%) patients reported old strokes, 39 (22.4%) had TIA, 27 (15.5%) had HIV-related cerebral infection, 5 (2.9%) had trauma, 8 (4.6%) had hypoglycemia, 5 (2.9%) had cerebral tumors, 14 (8%) were nonresidential Maputo patients, and 36 (20.7%) were excluded for other reasons. Overall, therefore, 651 (78.9%) of 825 patients had a clinically confirmed stroke-related event and were formally included in the study. Of these, 531 (81.6%) were classified as first-ever/de novo stroke events. Only 50 (17.7%) patients were clinically diagnosed without a CT scan or necropsy, 543 (83.4%) were confirmed by CT scan, and 58 (8.9%) were confirmed by necropsy examination. From these, 8 (1.3%) subarachnoid hemorrhages, 242 (40.3%) hemorrhagic events, and 351 (58.4%) ischemic events were identified. Of clinical and epidemiologic significance, more ischemic (42% versus 13.4%) and hemorrhagic events occurred as first-ever stroke presentations (36.1% versus 4.2%).
Figure 11.1 Incidence, exclusion, diagnosis, and type of stroke events.
As shown in Table 11.1, there were similar numbers of men and women, and the majority were of African ancestry (94.5%). Significantly, 12.9% of the stroke-related events occurred in patients aged <45 years. Those who arrived at the hospital on the same day as their onset of stroke symptoms represented just over half of the patients (58.4%). At the other end of the spectrum, 3.5% arrived >7 days after symptom onset. Overall, 177 (27.2%) and 171 (26.3%) of those with ischemic and hemorrhagic events, respectively, presented at the hospital on the same day as symptom onset. Far more de novo stroke patients arrived at the hospital on the day of symptom onset compared to those with a recurrent event (43.9% versus 9.5%). With regard to risk factors, hypertension was noted in 91.2% of the patients while AF was present in 6.1%.
Table 11.1 Patients’ prehospital characteristics according to type of case presentation.
| First-Ever Events | Recurrent Events | All Events (n = 651) | |||
| Ischemic (n = 264) | Hemorrhagic (n = 217) | Ischemic (n = 87) | Hemorrhagic (n = 25) | ||
| Women | 124 (47.0%) | 96 (44.2%) | 47 (54.0%) | 15 (60.0%) | 309 (47.5%) |
| Age (years) | 60.5 ± 13.8 | 54.7 ± 11.5 | 63.9 ± 11.8 | 57.2 ± 11.8 | 59.1 ± 13.2 |
| African ancestry | 249 (94.3%) | 212 (97.7%) | 77 (88.5%) | 23 (92.0%) | 615 (94.5%) |
| Education | |||||
| None | 77 (29.2%) | 55 (25.4%) | 27 (31.1%) | 7 (28.0%) | 182 (28.0%) |
| 1–4 years | 90 (34.1%) | 68 (31.3%) | 31 (35.6%) | 9 (36.0%) | 218 (33.5%) |
| 5 years | 49 (18.6%) | 35 (16.1%) | 9 (10.3%) | 5 (20.0%) | 104 (16.0%) |
| Time to stroke presentation | |||||
| Same day | 131 (49.6%) | 155 (71.4%) | 46 (52.9%) | 16 (64.0%) | 380 (58.4%) |
| 1 day poststroke | 55 (20.8%) | 45 (20.7%) | 21 (24.1%) | 6 (24.0%) | 140 (21.5%) |
| 2–7 days poststroke | 63 (23.9%) | 13 (6.0%) | 17 (19.5%) | 3 (12.0%) | 108 (16.6%) |
| Prestroke disability | |||||
| 0–2 (minimal) | 253 (97.4%) | 215 (99.5%) | 68 (78.2%) | 21 (84.0%) | 618 (94.9%) |
| Risk factors | |||||
| Hypertension | 213 (86.6%) | 196 (95.6%) | 80 (93%) | 24 (96.0%) | 561 (91.2%) |
| Cigarette smoking | 27 (10.5%) | 18 (8.5%) | 4 (4.8%) | 3 (12.0%) | 62 (9.8%) |
| Dyslipidemia | 37 (15.4%) | 23 (12.3%) | 11 (13.9%) | 4 (17.4%) | 82 (14.3%) |
| Diabetes | 40 (15.8%) | 20 (9.8%) | 16 (19.3%) | 3 (12.0%) | 87 (14.1%) |
| AF | 16 (6.2%) | 3 (1.4%) | 13 (15.7%) | 1 (4.4%) | 38 (6.1%) |
As shown in Table 11.2, case-fatality rates were high with a differential pattern according to ischemic versus hemorrhagic strokes and de novo versus recurrent events. Overall, 254 (39%) patients died during the 28-day follow-up, hemorrhagic strokes were related to 2- to 3-fold higher case-fatality rate, and median in-hospital stay was 6 days for the entire cohort.
Table 11.2 Patients’ short-term stroke-related case-fatality.
| First-Ever Events % (95% CI) | Recurrent Events % (95% CI) | All Events (n = 651) | |||
| Ischemic (n = 264) | Hemorrhagic (n = 217) | Ischemic (n = 87) | Hemorrhagic (n = 25) | ||
| In-hospital mortality (%) | 17.4 (13.0–22.5) | 47.9 (41.1–54.8) | 24.1 (15.6–34.5) | 52.0 (31.3–72.2) | 33.3 (29.7–37.1) |
| 7-day case fatality (%) | 16.2 (11.9–22.2) | 52.9 (43.0–65.1) | 17.3 (10.2–29.2) | 31.9 (15.2–67.2) | 32.9 (28.4–38.0) |
| 14-day case fatality (%) | 26.0 (20.2–33.4) | 69.0 (57.0–83.5) | 36.5 (25.1–53.3) | 63.1 (35.5–100) | 47.7 (42.1–54.1) |
| 28-day case fatality (%) | 27.4 (21.4–35.1) | 72.3 (59.8–87.5) | 40.7 (28.2–58.7) | 70.8 (40.6–100) | 49.6 (43.8–56.2) |
11.1.5 Study interpretation
Overall, 148.7/100,000 residents in Maputo per year required medical care for stroke and for hemorrhagic events; this resulted in in-hospital mortality of 50%, with case fatality rising to 70% by 28 days. These data confirm the serious impact of stroke on a population with high levels of often undetected and untreated hypertension, and complement more recent population incidence data derived from Tanzania [5]. However, as discussed in more detail below, it is clear that the hospital event-rates for stroke arising from this population may be conservative given a number of influential clinical, socioeconomic, and cultural factors that require consideration.
Despite the above factors, this study revealed much higher rates of stroke than the previous STEPS Stroke Multicenter survey [61]. However, both the high stroke incidence rate and the high proportion of hemorrhagic stroke events in Maputo are largely supported by the high prevalence of hypertension in Mozambique [62]. In agreement with these data, a contemporary, single-center study of hospital stroke presentations in South Africa reported that just over one quarter of presentations were hemorrhagic (27%), with a mean age of 51 years among those of African ancestry [59]. Previous studies report similarly high stroke-related case-fatality rates [59,66,67]. Poor survival rates reflect the limited pharmacological, technical, and human resources available to administer optimal treatment options. In Maputo, stroke patients are cared for in a general medical ward instead of in stroke units, where only one CT scan device is available for 1 million inhabitants. Furthermore, these data highlight gaps in secondary prevention due to insufficient access to treatment and monitoring of health conditions postdiagnosis. These data also emphasize the limited number of institutions available to rehabilitate individuals if they survive a primary stroke event. Physical rehabilitation therapy was offered to the patients free of charge; however, patients were required to fund their own transportation, restricting access to the minority with sufficient means to bear that cost. Overall, these findings are in agreement with previously reported results that the population of Mozambique is at a higher risk of adult mortality compared to other countries [68].
11.1.6 Study limitations
Importantly, these figures do not take into account additional stroke cases among those not residing in Maputo for more than 12 months. A small in-hospital stay fee was charged (as per the health system in Mozambique), but individuals with limited means and/or resources did not pay. Accessing hospital care with limited transport options is always challenging, and so these data remain conservative in terms of estimating the number and impact of stroke cases in this setting. The possibility of patients seeking health care in hospitals outside Maputo is highly unlikely as there are limited hospitals with no emergency medicine available to serve a population 10% larger and spreading over an area 87-fold larger than Maputo City [69,70]. Alternatively, owing to African culture (regarding stroke as a supernatural force and therefore requiring attention by a traditional healer [71]), local customs may have limited the number of stroke presentations to participating hospitals. From a clinical perspective, small hemorrhagic events may also have been missed in patients scanned long after the stroke event; however, the proportion of CT scans performed >7 days after stroke was relatively low and thus is unlikely to meaningfully influence the research estimates.
11.1.7 Study conclusions
In conclusion, the study’s surveillance approach strategy revealed a high incidence of stroke and related morbidity and mortality in Maputo, Mozambique. Given the rigorous methods used to derive such data, there is an opportunity to compare and contrast stroke-related events both within (on a historical basis) and without (in comparison to similarly derived data from other populations) the target population. Overall, these data reveal the deadly impact of stroke in Mozambique and wider sub-Saharan Africa, highlighting important issues around future stroke prevention (hypertension being a major target), life-saving treatments, and health services, as well as longer-term rehabilitation programs to mitigate the burden in vulnerable populations.
S5.2 Beyond counting cases: The next steps to improve health outcomes in stroke and beyond
After describing a troubling epidemiologic and clinical picture of stroke and indeed ACS, it is appropriate to outline some measures that could improve the situation. Stroke mortality reflects the control rate of hypertension in the community. Collectively, we need first to increase the awareness levels of hypertension in Africa to further increase its control rate. In countries such as Seychelles where this was achieved, researchers have already observed a decrease in the mortality rate of stroke [72]. The second measure is the creation of small and simple stroke units in the main hospitals where stroke patients can be admitted and undergo simple but efficacious procedures that are life saving. A small and simple unit with standard protocols and with personnel prepared for and dedicated to this type of patient has proved to be economically effective [73]. The same considerations around primary prevention and the need to develop a regional/national network of clinics to deal with an increasing and deadly caseload of ACS in younger African men and women may well emerge in the next decade or so. As discussed in Chapter 6, when combined with stroke events, there is a real potential for AMI and other forms of ACS to become the predominant forms of noncommunicable heart disease in sub-Saharan Africa, with more women and younger individuals affected relative to high-income countries.