The public, patients and doctors need to have better understanding of risk and uncertainty in health and healthcare. This includes the nature of the risks, in general, and the statistical probability of specific side-effects in an individual’s thera

RISK is part and parcel of everyday life. Notwithstanding medicine’s successes in the past half-century, there are both risks and uncertainties in the healthcare setting. Patients, doctors, hospitals, clinics, and medical suppliers are all exposed to a variety of risks and uncertainties. This article provides a perspective on health and healthcare risks.


Risk is defined in the Oxford dictionary as “the possibility of meeting danger or suffering, harm, loss, etc.” Within the context of healthcare, this specifically means the possibility of hazard, which is a set of circumstances that may have harmful consequences, including unfavourable outcomes for patients. For example, a bottle of medicines is a hazard if it is accessible to a child, exposing the child to a risk of harm. If, however, the bottle is locked in a cupboard and is thus, inaccessible to the child, the risk to the child is zero.

Measuring risks

There are various epidemiological methods used to work out risks. They include:

·Case reports, which may be the first sign of an adverse effect of a health condition or an intervention but by themselves, they provide a small idea of the magnitude of the risk;

·Cohort studies, which involve following up a large number of exposed people for a long period of time;

·Case control studies, which involve matching exposed people with controls who are otherwise similar. Although case control studies provide the most accurate measurement of risk, it is important to remember that perfect matches are not always possible; and

·Observational studies, which involve studying large numbers of people. This is particularly so when the risk is low.

There are also situations where it is not possible to carry out studies to measure the risks because of ethical considerations, e.g. the prescription of drugs during pregnancy. There are few drugs which risks during pregnancy are well documented. The safety of all the other drugs during pregnancy is not proven. In the case of a few of them, all that can be stated is that they have been used for many years without adverse effects.

Risk can also arise when the benefit of intervention(s) are not provided, e.g. taking folic acid in pregnancy. Hence, there are risks of both omission and commission.

The current A(H1N1) influenza pandemic has posed enormous challenges to frontline doctors who not only have to treat patients with influenza-like illness but also to assuage the “worried well”, and in the process, cause strains on the delivery system. – AFP

Absolute and relative risks

Risk is considered in either absolute or relative terms. It is vital that patients and their relatives understand both the absolute and relative risks of different health conditions and their treatment options. This should include the option of doing nothing.

Absolute risks provide information of how common an event is. However, relative risks provide an indication of the magnitude of the increase in end points but without reference to absolute risk.

A small increase in relative risk for a disease event, which occurs very frequently, will result in a large increase in the number of patients affected (large absolute risk).

On the other hand, a large increase in relative risk for a disease event, which occurs infrequently, will result in only a tiny increase in the number of patients affected (small absolute risk). When considering relative risk, one must never forget to base it on the magnitude of the absolute risk.

For example, a common condition may affect 10% of the population i.e. 100 out of 1,000 will be affected (absolute risk). If the relative risk is increased by 5%, the number affected will then be 105 out of 1,000 people. An uncommon condition may affect 1% of the population i.e. 10 out of 1,000 will be affected (absolute risk). If the relative risk is increased by 100%, the number affected will then be 20 out of 1,000 people.

This example illustrates the need to consider both absolute and relative risks when interpreting studies. In short, both the numerator and denominator must be considered, particularly the denominator.

When there are comparisons of risk, like have to be compared with like, as much as is possible, and not otherwise. Apples cannot be compared with oranges.

Risk implies that something is known about the balance of probabilities, particularly between the beneficial and harmful effects of medical interventions as well as health. Although the balance of probabilities may be expressed in mathematical terms, it cannot specify the outcome for a particular individual. In general, the interplay of several variables will determine the risk for any individual.

Risk presentation

Risk is presented in a variety of ways. Absolute and relative risks provide the information in numbers. Whenever relative risks are presented, it is crucial to consider the absolute risks to maintain perspective.

Other ways of risk presentation in medical journals, e.g. number needed to treat, odds and odds ratio, hazard rates, and ratios, etc will not be discussed here.

There are other ways of risk presentations for the public who are not keen on numbers. They include:

·Graphs, which present the numbers in a pictorial form.

·Comparisons of one risk to other risks, e.g. the increased risk of dying from smoking one cigarette each year is similar to the risk of being struck by lightning twice.

·Descriptions of risks as “high”, “low”, “common”, etc. Although, these descriptions are easily understood, they are not precise and mean different things to different people.

The Royal College of Anaesthetists of the United Kingdom have linked words to numbers when describing risks, i.e. very common means 1 in 10, common 1 in 100, uncommon 1 in 1,000, rare 1 in 10,000 and very rare 1 in 100,000 persons.

These numbers have also been put in a pictorial form for better understanding.

Risk perception

Risk messages, in most situations, are initially perceived by the source and not its content. The questions raised are who is the source and whether he or she can be trusted. If the answer to the latter is in the negative, the message is likely to be disregarded although it may be well presented. There is evidence that messages from distrusted sources lead to negative effects.

Although the content of the message influences risk perception, the manner of delivery particularly the emotional tone is crucial. Assessments of risk messages involve judgements about trust. Such judgements depend on what is done and what is said.

In fact, actions often speak louder than words. For example, the response to the message that masks need not be used by the healthy people in the prevention of A(H1N1) influenza pandemic would be influenced by the wearing of masks by the messengers during the presentation.

Trust grows when there is openness and willingness to listen. There is strong evidence that if trust is lost, its re-establishment is difficult and tedious.

Risk perception is also influenced by the potential for harm, the degree of familiarity with the situation, and the degree of uncontrollability of the risk. Certain factors elicit more anxiety and alarm than others. These so-called “fright factors” provide indications of which risks are perceived as worrying, or even alarming.

Risks raise concern if they are perceived as involuntary, e.g. exposure to cough droplets rather than voluntary, e.g. smoking; unavoidable despite taking personal precautions; arising from unfamiliar or novel source, especially if poorly understood by science or medicine; resulting from man-made rather than natural sources; resulting in awful illness, injury, or death; or posing danger or damage to pregnant women, small children, or to future generations.

In particular, risk perception and responses are aggravated by contradictory statements from authoritative sources, or worse still, from the same source. It is generally accepted that the more uncertain we are, the more afraid we are.

The responses to risk are influenced by people’s beliefs and values, which vary considerably. As most people value the motorcar, they accept the health risks of driving. Health messages on the dangers of drug abuse or misuse have limited impact on those who like to take risk.

The responses to risk are considerably influenced by the perceived benefits, or the lack of it. These benefits may be tangible or intangible.

How risk information is presented affects understanding of the risk and the response. When there is emphasis on one aspect of a health decision without including other aspects, there is impact on the understanding and perceptions of the risks and benefits.

Health information can be presented as positive or negative. For example, a patient can be informed that the oral contraceptive pill provides effective contraception in 99 of 100 (positive) or its failure rate is less than 1 in 100 (negative) when used according to advice. It can also be presented as a gain or loss, e.g. emphasis on the benefits of blood transfusion in an operation (gain) or the risks of blood transfusion (loss).

In general, there is more receptivity to health information presented positively and as a gain.

The responses to the A(H1N1) influenza pandemic have ranged from confusion, anxiety, and near panic to complacency and irresponsibility in response to the varying health messages. One message was that although A(H1N1) was a novel virus, most of those infected recovered from the infection and that personal hygiene was vital in controlling its spread.

Another message, which was recurrent, focused on the deaths. This was aggravated when wrong numbers were used in calculating the case fatality rate.

Yet another message was that the virus is an unseen enemy that spread rapidly to the community and there was much uncertainty about its propensity for mutation. There was different emphasis on these themes by different opinion leaders, few of whom were medically qualified, and some by the same opinion leaders.

The local situation was summarised very well by a speaker in the National Flu Conference who asked a pertinent question i.e. whether the A(H1N1) influenza is “a flu like any other flu or a flu like no other flu”.

In short, risk perception and its response is subjective, i.e. one man’s meat is another man’s poison.

Media influence

The print and electronic media, whether mainstream or otherwise, exercise a major influence on everyday life. The manner of risk presentations often influences its perception by the public.

There have been several occasions when doctors had to attempt to correct the information patients obtained elsewhere. These include the Nipah virus scare in 1998 and the oral contraceptive pill scare a few years earlier.

In the case of the former, the initial misdiagnosis resulted in human deaths and the near decimation of the pig industry. In the case of the latter, unwanted pregnancy rates increased markedly. Although subsequent studies did not confirm the risk, the distress caused to women was unimaginable.

The individual doctor’s responsibility of putting risk into context during such scares is almost impossible.

The current A(H1N1) influenza pandemic has posed enormous challenges to front line doctors who not only have to treat patients with influenza-like illness but also to assuage the “worried well”, and in the process, causing strains on the delivery system. Some doctors and other healthcare professionals have even become a part of the statistics.

Almost all doctors believe that more balanced reporting would have helped them in their very challenging tasks of treating those with influenza-like illness and allaying the anxieties of the “worried well”.

This is not to say that the media get things wrong or that they have been willful in exaggerating risks. Russell in an article, Living can be hazardous to your health: how the news media cover cancer risks published in a Journal of the National Cancer Institute Monograph in 1999, reported instances when some have been found to have done so with the collusion of some medical journals.

Going forward

The challenge for doctors is how to apply research evidence to individual patients in the light of the inherent uncertainty of medical evidence. For example, the most significant risk factor for lung cancer is smoking, which is responsible for 90% of all lung cancers. The risk is greater the more cigarettes are smoked. About one in seven long-term smokers get lung cancer. However, it is uncertain how it will affect an individual smoker, as there is, as yet, no means of detecting individual smokers who will get lung cancer.

Medical treatments often involve interventions with side-effects, which are unlikely to be totally hazard free. A balance has to be struck between potential benefit and potential harm. In principle, patients would not be exposed to risks, which cannot be justified in terms of a personal benefit from the treatment.

Although there are, as yet, no best practices for the communication of risks of harms and benefits, there are generally accepted principles of risk communication to individuals. They include informing patients of their risks, taking into consideration their medical, social and educational situation, transmitting current opinion, evaluating patients’ understanding of the discussion, listening to their concerns, and maintaining trust.

In respect of outbreak communication, the World Health Organization emphasise five critical factors i.e. building trust, early announcement, transparency, respect for public concern, and advanced planning.

The public, patients, and doctors need to have better understanding of risk and uncertainty in health and healthcare. This includes the nature of the risk, in general, and the statistical probability of specific side-effects in an individual’s therapeutic setting, in particular.

This is only possible if the patient has a regular doctor, as trust is the basis of the patient-doctor relationship. The development of the relationship, as in any other relationship, requires an expenditure of time and effort.

Patients have a legitimate right to competent care and minimal risks of treatment. They also have a right to expect that healthcare facilities are as safe as is possible. As such, regulators and the management of healthcare facilities have to ensure that effective and functional risk management procedures are in place, particularly in hospitals where the risks are much higher.

Minimising risks requires the participation of all stakeholders. It is not the responsibility of the doctor alone, as is often perceived. The patient’s role is, in fact, critical. A dynamic partnership between patients, regulators, media, medical suppliers, allied health professionals, and doctors will be crucial to making headway in minimising risks in healthcare.

Source: Dr Milton Lum