CHAPTER 14
Chronic heart failure

Okechukwu Samuel Ogah¹, Anastase Dzudie², Dike Bevis Ojji³, and Mahmoud Sani⁴

¹ University College Hospital, Ibadan, Oy, Nigeria

² Hospital General de Douala, Douala, Littoral, Cameroon

³ University of Abuja Teaching Hospital, Abuja, Nigeria

⁴ Bayero University Kano/Aminu Kano Teaching Hospital, Kano, Nigeria

14.0 Introduction

While there is a growing body of literature focusing on acute presentations of HF in sub-Saharan Africa, there is, perhaps understandably, a lesser focus on chronic HF, regarding both long-term management and outcomes. This chapter focuses on two seminal reports (one from Cameroon and the other from Nigeria) that provide important insights into not only the prevalence of HF in high-risk patient populations (unfortunately, population data are limited) but also the economic burden of HF. It is only with such knowledge that blueprints for the cost-effective primary and secondary prevention of HF in the unique African context can be applied.

14.1 Chronic HF among adults treated for hypertension in a cardiac referral hospital in Cameroon

Dzudie A, Kengne, Mbahe S, Menanga A, Kenflack M, Kingue S. Chronic heart failure, selected risk factors and comorbidities among adults treated for hypertension in a cardiac referral hospital in Cameroon. European Journal of Heart Failure 2008; 10:367–372. [98]

14.1.1 Study background

As noted in Chapter 13, reflecting its disproportionate impact in the developing world [99], hypertension is the most common form of noncommunicable disease in many African countries, and Cameroon is no exception [100,101]. A clustering of risk factors including type 2 diabetes, cigarette smoking, and obesity among those of African ancestry increases the probability of end-organ damage (i.e., renal disease, stroke, and HF) [102]. Although there is some (limited) information on the link between hypertension and stroke in the African context (see Chapter 11) [103,104], at the time of this study there were limited equivalent data on the development of chronic HF (noting the distinction between the acute manifestations of the syndrome outlined in Chapter 12); this was particularly true in Cameroon.

14.1.2 Study aims

The primary aim of this study was to examine the prevalence and characteristics of chronic HF among patients captured by the Hypertension Register from the outpatient department of the Yaounde General Hospital in Cameroon.

14.1.3 Study methods

14.1.3.1 Patient enrollment

Between 1995 and 2005, the Yaounde General Hospital was the main referral center for CVD cases in Cameroon, comprising patients from the capital city Yaounde and the wider country (population of approximately 15 million people). The hospital provides comprehensive, tertiary level facilities and management for high-level cases. Note that without a social security system, the cost of medical care was the responsibility of the patients and their relatives. As noted, this study focused on patients registered in the hypertension clinic of the hospital (typically attending a consultation twice a year, at which their medications were reviewed). Each patient was clinically profiled by trained nurses annually where possible. This included standardized BP measurements, 12-lead ECG, renal function assessment, fasting blood glucose, and serum electrolytes. Echocardiography was also performed based on clinical presentation. Clinical data were stored in a formal register. As part of the study, patients receiving treatment for hypertension in the clinic between February 1995 and January 2005 were evaluated for evidence of HF. The clinical notes of those with HF were reviewed for additional details, including sociodemographic profile and multimorbidity. Of 1,218 hypertensive patients registered during this period, 151 (12.4%) adult patients were identified as presenting with clinical HF or asymptomatic LV dysfunction; 4 (2.6%) patients with CHD were subsequently excluded in addition to 3 (2%) patients with VHD and 4 (2.6%) with missing echocardiographic data. Complete data were therefore available for 140 (92.7%) patients. This study received approval from the relevant ethics committee.

14.1.4 Study findings

14.1.4.1 Demographic and clinical characteristics

The overall prevalence of chronic HF in this hypertensive cohort of patients was 11.5%—rising from 7.2% in the first two years of the study to 12.4% in the final two years. As depicted in Table 14.1, HF patients were relatively young (mean age 55 years) with a predominance of men. Overall, 62 (44.2%) patients were NYHA Class III–IV, and 70 (50%) displayed echocardiographic evidence of LVH. The main form of HF was HFrEF (64%), with a residual 14% exhibiting asymptomatic LV systolic dysfunction. Isolated diastolic dysfunction occurred in 23% of patients. Overall, there were minimal differences between men and women, although significantly more women (p = 0.05) displayed isolated diastolic dysfunction (i.e. HFpEF). There were also minimal differences in BP levels and renal dysfunction between those with HFrEF and those with HFpEF (diastolic dysfunction).

Table 14.1 Patients’ demographic and clinical characteristics according to sex.

	Women (n = 54)	Men (n = 86)	All (n = 140)
Demographic Characteristics
Age (years)	52.5 ± 10.6	58.7 ± 14.6	54.9 ± 12.6
Risk Factors
Current Smoker	2 (3.7%)	16 (18.6%)	18 (12.9)
Dyslipidaemia	11 (20.4%)	10 (11.6%)	21 (15.0%)
BMI (kg/m²)	26.3 ± 4.1	23.3 ± 5.1	24.7 ± 4.7
Obesity	7 (13.0%)	9 (10.5%)	16 (11.4%)
Systolic BP (mm Hg)	152 ± 32	177 ± 41	164 ± 21
Diastolic BP (mm Hg)	93 ± 23	102 ± 15	98 ± 14.6
Controlled Hypertension	15 (27.8%)	17 (19.8%)	32 (22.9%)
Severe Hypertension	27 (50.0%)	31 (36.1%)	58 (41.4%)
NYHA Class
I - II	37 (68.5%)	41 (47.7%)	78 (55.7%)
III - IV	29 (53.7%)	33 (38.4%)	62 (44.3%)
HF profile
All systolic HF	38 (70.4%)	52 (60.5%)	90 (64.3%)
Asymptomatic LV Dysfunction	7 (13.0%)	13 (15.1%)	20 (14.3%)
Isolated Diastolic HF	19 (35.2%)	13 (15.1%)	32 (22.9%)
LVH	39 (72.2%)	31 (36.1%)	70 (50.0%)

14.1.4.2 Pharmacological management

Most patients were prescribed a combination of a diuretic (87.1%), ACE inhibitor/angiotensin II receptor blockers (68.6%), and/or calcium channel blocker (48.6%); beta blockers (11.1%) and centrally acting BP-lowering drugs (4.1%) were less common. Mean BP was 165/98 mm Hg and patients had been treated for an average of 7 years. At the last evaluation, 22.8% of patients had optimal BP control.

14.1.4.3 Risk factors

There were no sex differences in BMI profiles (11% obese), and dyslipidemia was evident in 15% of patients, with similar profiles in men and women. Smoking was more prevalent in men (18.6% versus 9.3%, p = 0.007).

14.1.4.4 Comorbidities

Figure 14.1 shows the distribution of comorbidities according to sex. No sex-related differences were observed in this regard; however, there were high levels of renal dysfunction and anemia among women and concurrent disease states such as chronic obstructive pulmonary disease (COPD) and arthritis, with minimal levels (just over 5%) of underlying CAD.

Figure 14.1 Pattern of concurrent morbidity in the study cohort according to sex.

14.1.5 Study interpretation

These data provide important insights into the prevalence of chronic HF in a hypertensive outpatient cohort in sub-Saharan Africa, with around 12% of individuals affected by the syndrome. Affected individuals were relatively young (middle-aged) and more likely to be male, with a clustering of risk factors and multimorbidity. Beyond the data presented in other parts of this section, these findings are particularly notable for exploring the prevalence of HF and asymptomatic LV dysfunction in patients with treated hypertension. The 11.5% prevalence of HF reported in this cohort is consistent with the 10.5% found by another Nigerian study [105], and both are likely to be an underestimate of the true prevalence of HF in this patient population. Isolated diastolic dysfunction in this cohort was similar to contemporary Caucasian population-based cohort studies [106] but higher than that (12.6%) described in another Nigerian series [107]. Renal dysfunction was revealed in one quarter of cases, representing a classic finding for this kind of cohort. This high frequency of renal dysfunction among a relatively young population is worthy of comment, with potential contributing factors including medication, low cardiac output, and the severe nature of hypertension among individuals of African ancestry. Numerous studies have demonstrated this to be an independent predictor of subsequent morbidity and mortality risk in patients with HF [108,109]. Diabetes, also evident but not highly prevalent in this cohort, likewise conveys an increased risk of poor outcomes; its prevalence being similar to that described in a hospitalized series of patients in Senegal [34]. Other noncardiovascular conditions such as COPD are progressively being recognized as important contributors to HF outcomes, and this may be particularly significant in the context of RHF (see Chapter 15) [110]. As reported in the Heart of Soweto Study [111], AF (found in 13% of this relatively young population) is becoming increasingly important in the context of HF, and given the age profile of the cohort, its prevalence and potential impact should be regarded as being relatively high.

14.1.6 Study limitations

Due to the retrospective nature of the study and the characteristics of the study population (which are likely to favor patients self-selected by symptoms from the total hypertension population), these data may provide imprecise estimates of the prevalence of hypertensive heart failure in the target population. As such, these factors may limit the generalizability of study findings.

14.1.7 Study conclusions

In conclusion, these data provide important insights into the prevalence of chronic HF in a particularly at-risk patient cohort. As noted in the original report, given the high prevalence of hypertension and other risk factors for HF in the Cameroon population at large [112], these findings are indicative of a future increase in the incidence of major target organ damages such as HF, kidney impairment, and stroke (often in the same individual). Primary and secondary prevention strategies alike are required to mitigate the future burden of chronic HF in Cameroon and wider sub-Saharan Africa.

14.2 The Economic burden of HF in Nigeria: Insights from the Abeokuta HF Registry Cohort

Ogah OS, Stewart S, Onwujekwe OE, Falase AO, Adebayo SO, Olunuga T, Sliwa K. Economic burden of heart failure: investigating outpatient and inpatient costs in Abeokuta, Southwest Nigeria. PLoS One 2014; 9(11):e113032. [113]

14.2.1 Study background

Until recently, little was known about the emerging problem of noncommunicable forms of HF supplementing traditional pathways to the syndrome in sub-Saharan Africa. A series of studies from South Africa [22,30] and Nigeria [16,87,97] (previously outlined in Chapter 6 and this section of the book) have added to historical reports on the syndrome by demonstrating very clearly that the etiology, natural history, and profile of chronic HF (i.e., more women and younger individuals affected in the prime of their life) diverge from high-income countries where the literature traditionally arises. It is now estimated that HF is responsible for 7%–10% of medical admissions in the region [16,97,114]. Significantly, given its potentially enormous cost implications (related both to direct health care costs and to the financial strain on affected individuals and their families), there is virtually no data on the economic burden of HF in sub-Saharan African countries and major populaces such as Nigeria.

14.2.2 Study aims

Given the paucity of data to describe the cost of HF both in its acute and chronic forms, this study was designed to determine the scope and cost of health care resources directed toward the treatment of HF in Abeokuta, Nigeria. Estimates were derived from a large and representative cohort of affected cases. Data were extrapolated to estimate the annual cost of HF in the region to inform health care policy toward efficient use of resources to mitigate the individual and societal impact of the syndrome.

14.2.3 Study methods

14.2.3.1 Patient enrollment

Study data were derived from the Federal Medical Centre, Abeokuta, Ogun State, southwest Nigeria. The Federal Medical Centre Abeokuta is a tertiary referral institution that receives patients from primary and secondary health facilities within and outside the state. All managed patients are given monthly appointments for clinical review as well as for refill of their medications. Patients pay out of pocket, and many are unable to buy medication that will cover longer periods. Significantly, patients are often cared for by their relatives, and health care costs in the city of Abeokuta and in most parts of Nigeria are typically borne by the patient as out-of-pocket expenses, given the limited access within the Nigerian population to social health insurance [115,116]. However, there are typically strong family and community bonds to care for the sick.

Data were derived from the Abeokuta HF registry. As described in more detail in Chapter 13, this was a hospital-based, prospective, observational registry that ran from January 2009 to December 2010, established to determine the current clinical profile of HF in the city as well as to assess typical clinical outcomes and health care costs (the focus of this report) associated with HF. In brief, all cases of HF presenting to the hospital were captured in the database. As noted in the original report, the study cohort was broadly representative of the population. During the study period, HF was responsible for 9% of total medical admissions [16], a figure that is consistent with equivalent reports from other parts of Africa [30,75,114].

14.2.3.2 Estimating health care expenditure

Costs were divided into two forms of activity: (a) direct health care costs associated with HF management and (b) indirect costs. The former comprised hospitals’ inpatient costs (e.g., cost of medical and nursing care, investigations, and pharmacotherapy). It also included the cost of surgical procedures and hospital transfer costs). Personnel costs included the opportunity cost of medical care, nursing, and ancillary support by other health workers (derived from Federal Government of Nigeria salary scales). For indirect costs, the days of lost work due to HF disability/illness (productivity loss) were calculated and the minimum wage used to monetize them.

14.2.3.3 Inpatient and outpatient costs

The standard costing table (unit costs) of the Federal Medical Centre, Abeokuta, for the year 2010 was used to compute the cost of consultations, hospital admissions, medical consumables, medical investigations, and pharmacotherapy. Outpatient costs were based on the costs associated (patients discharged alive) with monthly appointments for clinical review as well as refilled medications.

14.2.3.4 Specific costs

The cost of pharmacotherapy was based on the hospital’s price list (purchase price plus a 10% dispensing fee). The frequency of use of the various categories of drugs was based on the findings from the HF registry. The most frequent dosage was used for cost calculation. The cost of all laboratory and diagnostic tests was based on the price list of the hospital in 2010. The rate of consumption of these items was also garnered from the HF registry. The costs of surgeries and procedures were obtained from the most active surgical institutions in the region and were multiplied by the frequency of those surgeries and procedures.

14.2.3.5 Estimation of indirect costs

A human capital approach was applied with average annual earnings based on occupational group and average daily earnings then calculated per patient. The product of the working days lost and average daily earnings provided the productivity losses associated with HF in the study, and these were assigned monetary values.

14.2.3.6 Cost analyses

All available costs (as detailed above) were computed for the year 2010. An annual, prevalence-based approach was employed in estimating the cost of the resources used for the management of HF [117–120]. Health care costs were then expressed in the local currency—naira—and converted to US dollars (US$) (at a rate of 150 Nigerian naira to US$1 in 2010).

14.2.4 Study findings

Table 14.2 provides a summary of the clinical and demographic characteristics of the study cohort from which all HF-related health care activities and extrapolated costs were derived. The mean age was 58 years with just under half (46.9%) of the patients being female. Significantly, around one-third were aged ≥55 years and therefore in the prime of their potential working life.

Table 14.2 HF patients’ demographic characteristics in 2010.

	All (n = 239)
Demographic Characteristics
Women	112 (46.9%)
Age profile
Mean age (years)	58.0 ± 15.1
Aged 45–54 years	40 (16.7%)
Aged 55–64 years	59 (24.7%)
Aged 65–74 years	63 (26.4%)
Aged ≥75-years	33 (13.8%)
Married	168 (70.3%)
No formal education	88 (36.8%)
Unemployed	20 (8.4%)
Unskilled labour	146 (61.1%)
Resident of Abeokuta city	138 (57.7%)
Resident within Ogun State	53 (22.7%)
Resident outside Ogun State	48 (20.1%)
Monthly Income (Naira)
Unemployed*	18000 (US$120)
Unskilled labour	90,000 (US$600)
Skilled labour	240,000 (US$1600)
Professional	480,000 (US$3200)
Pensioner/Retiree	330,000 (US$2200)
Aetiology of HF
Hypertensive HF	195 (81.6%)
Dilated CMO	14 (5.9%)
Comorbidities
Hypertension	185 (77.4%)
Osteoarthritis	49 (20.5%)
AF	37 (15.5%)
Type 2 Diabetes	24 (10.0%)

^*Allocated minimum wage in the country.

14.2.4.1 Summary of inpatient and outpatient care cost

The total cost of inpatient care was estimated at N34,996,477 or US$301,230, comprising N17,899,977 (50.9%, $US114,600) and N17,806,500 (49.1%, $US118,710) for direct and indirect costs, respectively. Direct costs were responsible for 61% of inpatient care costs. Approximately 40% of the direct cost was due to surgery/procedures. Hospitalization, medical investigations, drug therapy, and transportation accounted for 20%, 24%, 15%, and 1% of costs, respectively. Outpatient care was estimated at N41,292,368 or US$275,282. Direct and indirect costs were N20,963,168 (US$139,754) and N20,329,200 (US$135,528), respectively, comprising 51% and 49% of total outpatient care costs. Transportation, medications, clinic visits, and medical investigations contributed 46%, 44%, 5%, and 5% respectively to these costs.

14.2.4.2 Total costs

The total estimated cost of health care attributable to HF in Abeokuta for the year 2010 was N76,288,845 or US$508,595. This translated to a cost of N319,200 or US$2,128 per patient per year. The proportional contribution of inpatient and outpatient costs was 46% and 54%, respectively. The contribution of various components to the total cost of HF to the region from an inpatient and outpatient perspective is summarized in Figure 14.2 and Figure 14.3, respectively.

Figure 14.2 Components of direct inpatient costs.

Figure 14.3 Components of direct outpatient costs.

14.12.5 Study interpretation

As described in the original report, this study represents the first systematic attempt to estimate the cost of HF in Nigeria (a major populace of sub-Saharan Africa) and the wider African continent. Costs were derived from an individual perspective, while the total cost per annum overall was calculated from a societal perspective. During the study period, 9% of total medical admissions were attributed to HF, a figure similar to earlier reports from other parts of Africa [30,75,114]. Overall, the total cost of HF or cost per patient per year (evenly distributed between inpatient and outpatient care) is enormous considering the context of a developing economy where out-of-pocket expenses are the main means of health care financing. A large proportion of estimated direct cost for inpatient care was attributable to surgical and medical procedures, while outpatient medications and transfers were responsible for 90% of the direct cost of outpatient care. This cost burden is not markedly dissimilar from reports derived from high-income countries [116,118,120,121]. Alternatively, the distribution of cost of HF in this setting is different from that observed in high income countries [117,118,121,122] but similar to one report from Brazil [120]. Significantly, there have been no reports from Africa with which to compare these findings. The contribution of hospital care cost in these countries ranged from 53% to 75%. Cost of outpatient care was in the range of 4% to 31% while the cost of drug therapy was between 6% and 8% of the total cost of care. The pattern of heart diseases as well as the level of technological development influence the mode of care, in addition to the utilization of sophisticated and expensive medical equipment, procedures, and consumables that are obviously needed for the care of HF patients. The consumption of these is higher in high-income countries than in this setting. This is clearly highlighted by the impact of surgery or procedures on the cost of hospital care. The few cases receiving surgery for valvular diseases or coronary interventions (out of the many that needed it) escalated the cost of hospital care in Nigeria. It is important to note that because of the younger age of affected individuals, HF in the Nigerian setting is associated with longer disability-adjusted life-years. By extension, this equates to a huge cost to the whole of society. With the changing demographic and epidemiological landscape in Nigeria coupled with the rising burden of cardiovascular risk factors and noncommunicable diseases (especially hypertension) in the country, the rate of HF is predicted to rise. Preventive measures are therefore critically needed but place an enormous strain on an already resource-poor health care system. The high cost of surgical interventions and procedures is out of the reach of the average Nigerian. Moreover, there is a need for a functional, effective, and efficient social health insurance system in Nigeria and other countries such as Cameroon (see Section 14.1 above). Unfortunately, most of those afflicted by HF are poor and are unlikely to sustain the treatment of their illness for an extended period. There is also a need to develop community-based HF care in Nigeria and other parts of sub-Saharan Africa, as this will reduce the cost of outpatient care, which is largely influenced by the substantial cost of frequent transport to and from the health facility for specialist HF management.

14.2.6 Study limitations

A few limitations are worth noting. First, as this study was conducted in a tertiary care facility, it is possible that mild cases may have been missed. Second, cost based on the severity of HF was not assessed, nor were age and sex-specific cost analysis. Third, the cost of comorbidities were not included, nor the cost due to alternative medicines, over-the-counter purchases, capital cost, and indirect cost by caregivers as well as the general cost to society.

14.2.7 Study conclusions

These seminal data, representing a first of a kind in the African context, show the profound impact and importance of HF as a major public health problem in a developing economy such as Nigeria. Unfortunately, the country is still battling with communicable diseases. As described in Soweto, South Africa (see Chapter 6), there is an obvious confluence or crossroads of communicable and noncommunicable disease in vulnerable populations with limited health care resources and personal wealth. The annual individual cost of HF is high, coupled with the fact that out-of-pocket expenses in Nigeria compose close to two-thirds of health care costs. There is urgent need, therefore, to reduce HF-related expenditure and cost through the control of known risk factors for the syndrome (see Section 3). As outlined in Chapter 13, this particularly applies to elevated BP levels in many rural and urban communities due to epidemiological transition and a movement away from traditional lifestyles that previously protected individuals from noncommunicable diseases. At the same time there is an imperative to reduce hospital as well as outpatient care costs through the development of community-based HF care programs that have the potential to reduce the high levels of recurrent hospital stay and case-fatality already seen in this population [123,124].