Introduction

Hypertension is the most important risk factor for cardiovascular diseases (CVD), accounting for about 54% of strokes, 47% of ischemic heart disease (IHD), 25% of other CVD, 13.5% of all deaths and 6% of all disability-adjusted life-years (DALYs) (the sum of years lived with disability and years of life lost) worldwide. [1] It is of note that about 80% of the burden occurred in low and middle-income countries, and a greater proportion of the burden was in young people of working age in low and middle-income regions than it was in high-income regions.[1] In a recent meta-analysis, the overall prevalence of hypertension in Nigeria was estimated at 28.9%, with a prevalence of 29.5% among men and 25.0% among women, and of 30.6% and 26.4% among urban and rural dwellers respectively.[2] In addition, the pooled awareness rate of hypertension in the country was low (17.4%).[2] These estimates hold true even among semi-nomadic Fulani rural population, among whom we recently reported a prevalence of hypertension of 28.5% in a rural settlement in Kano, North-West Nigeria.[3] In this Fulani population, 39.3% of the hypertensives were aware of it, and only 25% of the known hypertensives were on treatment. Age was the only independent predictor of hypertension, and for every increase in age by 1 year, the odds of developing hypertension was increased by 6.6%.[3]

Among hypertensive in-patients, Kolo et al retrospectively studied hypertension-related admissions and their outcome over 1 year, and reported that hypertension-related admissions represented 23.7% of the total, while mortality attributed to complications of hypertension was as high as 42.9% of the total.[4] Arodiwe et al similarly reviewed hypertension-related admissions over 5 years, and reported that hypertension-related admissions represented 6.2% of the total, with a case fatality rate of also 42.9%.[5]

The morbidities and mortality in hypertension impact negatively on many outcome measures among hypertensive subjects. Ukoh VA reported that the most common hypertensive complication was stroke followed by heart failure (HF) and chronic renal failure (RF), and 10.5% of all medical admissions was due to hypertensive complications.[6] Furthermore, Onwuchekwa et al reported that among admitted hypertensives, stroke was responsible for 39.9%, HF for 22% and RF for 9.4% of hypertensive complications, while 51.5% of deaths were due to stroke, 14.1% due to HF, and 12.1% due to RF.[7]

The above-mentioned studies were single centre based and of modest sizes; hence not representative.[5-7] It is therefore clear that comprehensive information is needed on the burden of morbidities, clinical characteristics and mortality related to systemic hypertension in a multicentre study, which can fill the knowledge gaps on hypertension in Nigeria. The present study therefore aimed to systematically assess the cardiovascular and other co-morbidities and in-hospital outcomes among hypertensives admitted to Teaching Hospitals in three large metropolitan cities within two geographic zones in Nigeria.

Methods

This was a retrospective study carried out in Departments of Medicine, Aminu Kano Teaching Hospital (AKTH), Kano, Ladoke Akintola University of Technology Teaching Hospital (LAUTECH TH), Ogbomoso, and Lagos University Teaching Hospital (LUTH), Lagos, Nigeria.

Approvals to carry out the study were sought from the Ethics Committees of the 3 study centres before the commencement of the study.

Medical records of all subjects admitted to the medical wards of the study centres with an established diagnosis of hypertension in 2013 were reviewed. Admission, discharge and mortality registers of the medical wards were used to identify the cases, those discharged and those who died. The records of the patients were then reviewed and included if the following inclusion criteria were satisfied: (1) established diagnosis of systemic hypertension, (2) availability of all admission notes, (3) availability of results of investigations that could confirm clinical diagnoses, and (4) documentation of required information for the study. Hypertension was defined as presence of sustained systolic and or diastolic blood pressure (SBP and DBP respectively) of ≥140 and or 90mmHg respectively, or its documented history, or if a patient was on any antihypertensive medication even if BP was normal.

The retrieved data for the study included demographic and clinical characteristics, clinical diagnoses, results of confirmatory and supporting investigations, duration of hospitalisation (in days) and information on whether patients had died or survived hospitalisation and were discharged or referred.

The data was explored for skewness and analysed using SPSS version 17.0 software. Proportions, medians with interquartile ranges and means with standard deviations were used to summarise patients’ characteristics, as appropriate. Chi-square, Fisher’s exact probability, Student’s t and Mann-Whitney tests were used to compare categorical and continuous variables, as appropriate. Regression models were used to determine predictors of mortality, and results were expressed as Odds ratio (OR) with 95% confidence intervals (95%CI). Survival analysis was conducted using Kaplan-Meier curves and Log rank test obtained for all-cause mortality between male and female patients. A p-value of <0.05 was considered statistically significant.

Results

In 2013, the number of beds in the medical wards of AKTH, LAUTH and LUTH were 68, 100 and 120 respectively. The admission records showed that a total of 288 hypertensive patients were consecutively admitted in the medical wards of the 3 centres in 2013, but 44 of them had incomplete medical records. The remaining 244 hypertensive patients satisfied the inclusion criteria and were included. Of these, 146 (59.8%) were males and 98 (40.2%) were females. The mean age of all patients was 58.4±15.9 years with a range of 20-93 years, mean SBP was 150±37mmHg and mean DBP was 94±24mmHg. In addition, 64 (26.2%) were unemployed, 154 (63.1%) were gainfully employed and the remaining 26 (10.7%) were retirees.

The baseline characteristics of males were compared with those of female subjects in Table 1, which shows that majority of males (67.1%) were urban dwellers and majority of females (60.2%) were rural dwellers, and males spent significantly more days on admission than females, with a median of 15 vs 9 days respectively. However, the mean age, blood pressures, heart rate and other variables in the Table were not significantly different between the 2 groups.

The pattern of antihypertensive prescriptions at admission is presented in Figure 1, which shows that 31.7% of the patients were not on any antihypertensive drug while the majority (62.3%) was on either 1 or 2 antihypertensive drugs. Further analysis showed that the prescription pattern did not differ by gender among the subjects (p=0.436).

Co-morbidities among male and female patients are presented in Table 2 and Figure 2. The commonest co-morbidity among all patients was HF followed by stroke/TIA, seen in 76 (31.2%) and 69 (28.3%) patients respectively. The most frequent co-morbidity among males was HF seen in 34.3% of them, while stroke and TIA were more common among female patients, seen in 34.7% of them. Stroke and TIA was also the 3rd most common among males while HF was the 2nd among females. In addition, excess alcohol, cigarettes smoking and peripheral artery disease were significantly more common among males, while dyslipidemia and DM were significantly more common among females. Atrial fibrillation was the commonest arrhythmia seen in 5 (3.4%) male and 1 (1%) female, junctional rhythm in 2 (1.4%) male and 2 (2%) female, atrial flutter in only 1 female (1%), supraventricular tachycardia in 1 male (1.7%) and 1 female (1%) patients, and ventricular tachycardia in only 1 (1%) female patient. Non-cardiovascular co-morbidities were uncommon, but the most frequent among both males and females was CAP seen in 7.4%, followed by chronic liver disease in 3.7% and neoplastic diseases in 2.9%, of all patients. Overall, 88.4% of males and 87.8% of females had 1 or more co-morbidities at admission, as illustrated in Figure 2.

Figure 1.

Antihypertensive prescriptions among male and female patients MNumber of patients on different combinations of antihypertensive drugs, according to gender.

A total of 19 patients (7.8%) comprising of 13 (8.9%) males and 6 (6.1%) females (p=0.427) died in-hospital while others were discharged or referred. The survival curves for male and female patients were constructed and data was censored at 50 days (Log Rank (Mantel-Cox) p= 0.970) (Figure 3). The characteristics of the deceased were compared with those of the discharged patients in Table 3, which shows significantly higher mean SBP and prevalence of those not on antihypertensive treatment at admission among the former than the latter group (p=0.045), while other variables were not significantly different between the 2 groups. In addition, prevalence of stroke tended to be higher among the deceased while that for HF and DM tended to be higher among the discharged patients, although the differences were not statistically significant. Further analysis showed that not being on antihypertensive treatment at admission increased the odds of mortality by 7.5 fold (OR=7.5; 95%CI=2.8-20.0; p<0.001) in the regression models.

Discussion

In the present multicentre retrospective study involving consecutively admitted hypertensive patients in 2013, data was complete for inclusion for only 244 patients. Two-thirds of the patients were on antihypertensive medications at admission, and males spent more days on the admission than females, with a median of 15 vs 9 days respectively. Various co-morbidities were found among 88.4% of males and 87.8% of females, and the most frequent among males was HF while stroke and TIA were most common among female patients. Arrhythmias were found in only 5.7%, and the most frequent was atrial fibrillation in 2.5% of the patients. Non-cardiovascular co-morbidities were uncommon, and the most frequent among all patients was CAP. The prevalence of in-hospital mortality was 8.9% among male and 6.1% among female patients (p=0.427), which was mainly predicted by not being on antihypertensive treatment at admission.

Firstly, this study has revealed the degree to which medical records were being kept in the study centres in 2013. The results show that 15.3% of the admitted hypertensive patients were not eligible for inclusion in the study because of incomplete records. Although this is significantly better than the observation made in a similar study where only 54.4% of the records had complete data for inclusion, our finding is still worrying.[7] These observations reveal the weaknesses of manual medical record keeping, and call for a shift towards computerised medical records keeping which was not available at all the study centres in 2013, which would have substantially reduced the challenges of missing or incomplete medical records. The unexpectedly low admission rates for patients including hypertensives in the study centres in 2013 could have been due to closures of the hospitals because of frequent strike actions by health workers.

Figure 2.

Pattern of co-morbidities among male and female patients Legend: Number of patients admitted with various comorbidities described according to gender.

Figure 3.

Kaplan-Meier curves of survival among male and female subjects.Legend: Kaplan-Meier curves of overall survival among hospitalised male (green line) and female (blue line) hypertensive subjects, constructed after excluding subjects who spent >50days on admission. Log Rank p=0.970

Secondly, the prevalence of co-morbidities was high in both males and females, and 88.4% of males and 87.8% of females had 1 or more co-morbidities at admission. The commonest co-morbidity among all patients was HF (31.2%) followed by stroke/TIA (28.3%); the former being commonest among males (34.3%) and the latter among females (34.7%). As mentioned earlier, systemic hypertension is the most important risk factor for CVD, associated with important morbidities and mortality.[1] It has also been reported that 75.6% of hypertensive patients in Nigeria tend to be at very high risk for CVD events even if on treatment.[8] As such, it is not surprising that over 80% of the admitted patients in the study centres had such high prevalence of cardiovascular morbidities, including HF, stroke, DM, dyslipidemia and renal failure. Previous studies in Nigeria have shown similar morbidity pattern among hypertensives, with stroke as the most common complication followed by HF, while ischemic heart disease (IHD) is rare.[4-7] The rarity of IHD could be related to the pattern of the other CVD risk factors among our patients, because high blood pressure, heavy alcohol use, and advanced age are stronger predictors for stroke than for IHD, whereas dyslipidemia, diabetes mellitus, and smoking are more strongly linked to IHD.[9] Among 192 World Health Organization member countries, stroke burden was disproportionately higher in China, Africa, and South America, whereas IHD burden was higher in the Middle East, North America, Australia, and much of Europe.[10]

It is of note that the non-CVD co-morbidities were uncommon among patients in the present study, and hence not likely to be the reasons for the hospitalisations. Our observations therefore suggest that hypertensive patients in Nigeria tend to suffer mostly from cardiovascular complications than from other illnesses.

The third important observation in the present study is that in spite of the high prevalence of co-morbidities of hypertension, including HF, stroke, DM and renal failure, the in-hospital mortality was relatively low (7.8%), without significant gender difference. In comparison, reported mortality rates among admitted hypertensive patients in other Teaching Hospitals in Nigeria were 34.3%, 23.3% and 22.1% respectively.[4,6,7] In agreement however, most deaths were consistently caused by stroke, followed by HF or RF.[4,6,7] The deceased had higher SBP than the survivors, and majority of them (52.6%) were not on any antihypertensive medications at admission, which was the only predictor of mortality in the present study, with odds of dying increased to 7.5. Unfortunately, data on treatment patterns was not provided in the 3 other studies from Nigeria for comparison with ours.[4,6,7] However, it is possible that the differences in treatment patterns between the studies could explain the variability in the mortality rates. It is well-documented that high BP bears an independent continuous relationship with the incidence of several CV events including stroke, myocardial infarction, sudden death, HF, peripheral artery disease and end-stage renal disease, at all ages and in all ethnic groups.[11] Clinical trials of antihypertensive therapies have demonstrated their benefits in terms of progressive reduction of death and morbidity from strokes as well as from coronary heart disease in the order of approximately 70% and 55%, respectively.[12]

Limitations

Incomplete medical record is one of the inherent limitations of retrospective studies, including the present study. In addition, data on the direct causes of deaths was not available to us and was not presented.

Conclusion

The present multicentre retrospective study involving consecutively admitted hypertensive patients has described the cardiovascular and other co-morbidities and in-hospital outcomes among hypertensives admitted to 3 Teaching Hospitals in 2 geographic zones in Nigeria. Two-thirds of the patients were on antihypertensive medications at admission, and co-morbidities were found among 88.4% of males and 87.8% of females. The most frequent co-morbidity among males was HF while stroke and TIA were most common among female patients. Non-cardiovascular co-morbidities were uncommon, and the most frequent among all patients was CAP. The prevalence of in-hospital mortality was relatively low, and not being on antihypertensive treatment at admission increased its odds by 7.5 fold.