The Social Impact of Norvatis medicines: case Studies from South Africa and Kenya

About the Company

Novartis is a Swiss multinational healthcare company that provides solutions to address the evolving needs of patients worldwide.  At the end of FY18, the organization recorded about 125,000 employees and $52 billion in net sales. Novartis applies science-based innovation to address some of society’s most challenging healthcare issues.

About the Approach

In the last few decades, healthcare expenditure has been increasing. This has been accompanied by a notable parallel increase in average longevity and quality of life among the global population. As a result, efficient spending in healthcare is increasingly recognized as a direct predictor of better health outcomes and national wealth. Thus, upfront national spending in health systems that is conditionally bound to bring about better quality of life and wellbeing upon the respective populations could be considered a form of national investment.

Measuring health-related quality of life precisely and reliably has been, nonetheless, a longstanding challenge in public health. Capturing and quantifying a universal unit of increase or decrease in quality of life on the individual and collective patient levels enables economists to monetize such unit into an economically comprehensible monetary outcome that is compatible with traditional validated economic research techniques. Significant strides in the last few decades were taken to address the conceptual and ethical challenges in this regard, resulting in an increase in the quantity and quality of the body of evidence being published.

Stage 1: Frame

Get Started

Social impact, both positive and negative, is a key element of Novartis´ Financial, Environmental and Social (FES) impact valuation, the Novartis version of the Triple Bottom Line approach. In this study, a novel framework that is generalizable and scalable across different countries and medicine portfolios has been developed. This framework capitalizes on a plethora of published medical literature that has been substantially growing in the last few decades. This literature allows for the estimation of health gains generated by Novartis’ medicines expressed in Quality Adjusted Life Years (QALYs) and Years of Life Saved (YLS). Through the subsequent steps of this approach, health gains are translated into gains in paid and unpaid work activities. Owing to a healthier and more active patient population, those gains eventually contribute “monetary revenue” to the national Gross Domestic Product (GDP). This monetary revenue is referred to as the Social Impact.

Stage 2: Scope

Define the objective

The aim of this study was to quantify and value the Social Impact of the entire Novartis global product portfolio in monetary terms.

Scope the assessment

The study is still ongoing with an expanding three-dimensional scope (time, geography and medicines covered), however 63 Novartis medicines sold across 11 countries from 3 different portfolios in 2016 and 2017 were assessed (see below).

Stage 3: Measure and Value

Measure impact drivers and/or dependencies

The study was conducted using a phased approach to calculate the global social impact of the Novartis medicines in scope. The first phase involved calculating the drugs´ Health Benefits through the incremental gains in QALYs and YLS on the relevant patients. These were then aggregated and translated into activity gains comprising paid and unpaid work. Last, the Socioeconomic Benefits were calculated as monetary contributions to the national GDP in US dollars.

Measure changes in the state of social & human capital

The Health Benefits

Comprehensive literature searches in MEDLINE (accessed through PubMed) and Google Scholar were conducted. The objective was to identify published economic evaluations quantifying QALYs as the utility/effectiveness measure for every medicine and sub-indication included in the study. QALYs/YLS were selected as they demonstrate health outcomes across diverse diseases. The incremental undiscounted QALY/YLS gains compared to the Standard of Care (SoC) were then calculated for the average patient for one year. For medicines with multiple authorization labels (clinical uses), the epidemiological weight based on prevalence estimates from the Global Burden of Disease (GBD) study was used. This facilitated deriving an average indication-weighted QALY estimate for an average patient receiving the medicine.

Furthermore, the proportion of patients in the working age (under 60 years of age) were also derived from the GBD study to later differentiate between economic gains from paid and unpaid work activities. Whenever literature reporting QALYs was not available, Years of Life Saved (YLS) were used as an alternative metric of health gains. Similarly, when QALYs and YLS were not found for a specific medicine, the Anatomical Therapeutic Chemical (ATC) classification system was used to derive QALYs/YLS for proxy medicines that were nearest in classification to the medicine in question.

On the other hand, when multiple suitable publications were available for one medicine and indication, selection of the best match was based on ten criteria that aimed at arriving at the best available evidence. Those criteria (listed below) prioritized (in descending priority), the literature providing an overall closer match to the country and disease indication of interest when comparing competing sources:

  1. Disease indication and medicine label
  2. Medicine comparator
  3. Patients’ demographic and disease characteristics at baseline
  4. Time horizon of study
  5. Country investigated in the study
  6. Medicine dose
  7. Medicine dosage form
  8. Discounting rate
  9. Publication date
  10. Health outcome investigated

Subsequently, the QALY estimates for every medicine were multiplied by the number of patients reached for the corresponding medicine sold in the country and year of interest. The figures on patients reached were provided by Novartis.

The Socioeconomic Benefits

In a second step, activity gains associated with improved health were quantified from a macroeconomic perspective. This was achieved through linking QALY gains with a measure of patient’s paid and unpaid work activities. Country specific parameters from macroeconomic databases by the International Labour Organization (ILO), United Nations (UN) or the World Bank were used.

To estimate a measure of paid work for individuals in the working age, gained QALYs were valuated against the average annual labour productivity, i.e. the country specific gross value added (GVA) per employee. Thus, it was assumed that all patients who are younger than 60 years of age are economically active (either on full or part time basis) and that no children were among the patients reached.

To quantify the activity gains beyond employment, information on the average time use in hours per day, reported separately for males and females, was used as a basis to attach a monetary value for unpaid work to each QALY. Data on unpaid work activities was only available in highly aggregated form for most of the country portfolio. On this account, the amount of unpaid work in terms of GDP contributions was approximated in two steps. First, built on the assumption that GDP per capita reflects the amount of paid work per capita, the measure was multiplied by the ratio of time use for paid and unpaid work per capita. This ratio can be interpreted as people spending a factor of the amount of time for paid work additionally on unpaid work. In a second step, the resulting product was multiplied by an estimated factor, which was intended to reflect that unpaid work activities have a lower labour productivity than average across all sectors of the economy (see figure below).

In addition, the wider economic indirect and induced effects initiated by an increase in economic activity were considered by using country-specific value-added multipliers.

For Kenya and South Africa, some special features had to be considered during implementation. In the case of Kenya, there was no data on time use available from the United Nations time use portal or any other source. Therefore, it was necessary to identify a country whose values could be used as best proxy. The choice of Tanzania as best representative was based on proximity of place as well as highest convergence in GDP per capita and Human Development Index (HDI).  Additionally, information from national accounts by economic sector were retrieved from the Kenya National Bureau of Statistics (KNBS), since other sources did not contain all the data needed. Data for South Africa was available. Thus, no proxies were needed to complete the parameter set.


Overall, 46 publications were used to derive QALY estimates for 33 medicines and YLS estimate for one medicine (amlodipine). Total patients reached per medicine varied widely, and therefore the number of patients reached per medicine and country along with the country specific economic parameters influenced the total Social Impact of a drug portfolio in the corresponding country.

For the year 2017 in South Africa and 2016 in Kenya, based on the numbers of patients reached, a total of 56,711 QALYs and 59 YLS were generated through the use of the 34 medicines. This amounted to a total of USD 1.95 billion as monetized Social Impact.

Epidemiological data were used to estimate the average proportion of patients in the working age for the target population of each medicine. The calculated Social Impact per medicine reflects the age structure of the underlying patient population when examining the proportional contribution of the two components of the Social Impact: The GVA due to paid and unpaid work activities.


Stage 4: Apply and Integrate

Interpret and Test The Results

WifOR, in close cooperation with Novartis, had a priori defined and well-agreed-upon study concept and methodological framework in place. Nevertheless, due to the novelty of this approach, this pilot study was conducted while having to frequently solve problems on an ad-hoc fashion.

The current report arguably delivers ground-breaking analyses that provide insights and visibility on value aspects for medicines on an unprecedented global scale. In this section we list the main set of assumptions that we expect to have impacted the certainty of our estimates. We believe that the uncertainty brought about by those assumptions are acceptable given the bird’s eye perspective this study intends to deliver and the explorative nature of the pilot project. The main assumptions of the study are the following:

  1. The health gains reported for a studied population in the literature did not always coincide perfectly with the target population of the country, drug and indication in question. The degree of precision to which a piece of literature depicted the health outcomes upon the intended target population depended on the extent of their resemblance with regards to the ten characteristics listed in the Measure & Value Stage.
  2. While QALY is an aggregate metric of survival and quality of life, we assumed that one QALY is equivalent to one person-year of full capability of performing paid and unpaid activities.
  3. Economic evaluations that only reported discounted QALYs/YLS were still included in the study. In the absence of the full economic models, information on the actual undiscounted QALY gains per year could not be precisely reproduced. We used the conventional discounting formula together with the reported discount rate and half the reported time horizon to derive an estimate of the undiscounted QALY. This method, however imperfect, was validated using economic evaluations reporting both discounted and undiscounted QALYs and was found to be conservative. The discounting formula, where P is the discounted value, F is the undiscounted value, r is the discount rate, and t is the time horizon.

  1. Except for information on the age structure derived from GBD study, further details on the patients’ characteristics were not pursued. It was assumed that an average patient shares the same economic profile of the population’s average person, e.g. the amount of time spent working.
  2. Expressing the Social Impact in USD (not the national currency) unties the relation between the monetary gains to the living costs in the respective country.
  3. For Kenya and South Africa: Since both are developing countries, it is expected that the informal economic sector plays an important role in those countries’ economies. Not taking the informal sector into account in our calculations might have biased our estimates when calculating the GVA generated due to unpaid work.

Take Action     

Reducing the hurdles for other companies to develop similar approaches is key for mainstreaming social and human capital measurement and management. Novartis uses various platforms to share its Financial, Environmental and Social impact valuation approach. The Embankment Project for Inclusive Capitalism recognized the approach in its report.