Statistics- the glass is half-filled or half-empty

Most medicos hate biostatistics but have to face it at one time or another with varying intensity. It collects, analyses, and interprets data providing evidence crucial for health policy and planning. Also, biostatistics is more often abused/ misused than used from the correct perspective. The same data set can build up different and diametrically opposite narratives. It is like a glass half filled with water. Some will say it is half empty while others say it is half-filled (both are correct). Several jokes are prevailing about this subject such as (1) Lie, damn lie, and statistics (in that order) or (2) Statistics is like a bikini that flaunts unnecessary details but conceals the crucial components.

Any health-related issue can be described in terms of absolute numbers or rates. While both are true, they can give different interpretations and produce different impacts. As such both absolute numbers and rates are important. Rates ensuring comparability help in ranking the groups/ territories in terms of their performance and absolute numbers show the magnitude of the problem (crucial for planning for manpower/ infrastructure & budgetary allocation). Let us look at a few of the examples.

1. Maternal Mortality Ratio (MMR):

MMR in India is 97 per 100,000 live births. This may not produce the desired impact on people as it is a small number but if we say that in India, 24000 women in their prime age die every year due to a cause that is fully preventable with simple interventions, is more impactful. It can further be dramatized by saying that every third day an airplane takes off and crashes killing all unlucky innocent passengers. Though the information provided is the same the second statement will have more impact. At the same time, if it is said that at the time of independence, our MMR was 2000 per 100000 live births, we can glorify (rightly so) our achievements by saying that we have reduced it by more than 95 percent. However, these achievements will look bleak if we are told that Scandinavian countries (Norway, Sweden, or Denmark) have MMR between 2-4 per 100,000 live births, and Ireland had zero maternal deaths in 2020.

2. Adverse sex ratio:

The sex ratio is expressed as the number of females per 1000 males. Normally with no intervention from outside, this ratio should slightly favoring females (> 1000). However, in India, the sex ratio for the total population has been 927, 933, and 940 (/1000 males) during the censuses of 1991, 2001, and 2011 respectively. The adverse sex ratio was for the first time noticed by the authorities which alarmed everyone and as a result, several actions aimed at preventing prenatal sex determination (followed by female feticide) and generating awareness and incentivizing for promoting female childbirths were initiated. These numbers appear small and did not alarm the general population but when these rates were applied to the population, it seems some 3 million female children were not allowed to be born between 1991 – 2001. The sex ratio between 0-6 years which reflects the fertility experiences of the last 6 years continued its decline from 945 (1991) to 927 (2001) to 914 (2011). Therefore between 2001 and 2011, things would not have been different. Thanks to Corona, we do not have census data for 2021. Therefore, talking in terms of the disappearance of female children in such a large number is more impactful in drawing the attention of everyone to the ill effects of this adverse sex ratio than talking in terms of the changing sex ratio by a few points this or that side.

3. Economic development:

We have a strong and emerging economy and there is every reason to cheer for it. Our economy of $ 1.0 trillion (2010) has increased to $ 3.89 trillion (2024).  Currently, it ranks as the 5^th largest economy on nominal and as the 3^rd largest economy on purchasing power parity (PPP) basis. India has the fastest-growing economy across the globe; growing @ 7% annually.  Another report from NITI Aayog, using the multi-dimensional poverty index (MPI), claims that some 250 million people have been brought out of poverty between 2013-14 and 2022-23. Accordingly, in this period, the proportion of the poor also decreased from 29.2% to 11.3%.

But if we look at the flip side and consider per capita GDP, it is $ 2698 (2024), ranking India at 129 out of 200 countries (nothing much to cheer). Further, considering the fallacy of mean, we come across economic inequality. According to the World Inequality Report 2022, India is among the most unequal countries in the world, with the top 10% and top 1% of the population holding 57% and 22% of the total national income respectively. United Nations Development Programme (UNDP) reported that India has the highest number of people (234 million) living in poverty. This report places India as home to the maximum number of poor people in the world. So, a narrative can be built up based on your leanings using the same data set.

4. Data management:

A downside of statistics is its nature to draw doubtful correlations which may lead to health planners making incorrect decisions.  During the coronavirus pandemic, the reduction in the number of cases in the later phase was attributed to the mass vaccination, whereas it could have been due to the development of herd immunity. It led to a more vigorous policy of compulsory vaccination. Finally, data management reminds me of the following famous quote.

“If you torture the data long enough, it will confess to anything”

(Ronald Coase, London School of Economics)

My final take in this regard is that any statistics need a cautious interpretation because it can create either type of picture depending on who is doing it and with what intention.