## R

The proportion of the total variation in the dependent variable of a multiple regression that is 'explained'. It is the squared correlation between actual and predicted values of the dependent variable. A measure of the goodness-of-fit of the equation. Also known as the 'coefficient of determination'.

The diagnostic use of radiation such as X-rays to make images. Radiology

The science of X-rays and high-energy radiation in imaging and treatment processes. Same as nuclear medicine.

The treatment of cancer by X-rays or gamma rays to destroy cancer cells. Random Effect Model

A model in which treatments are a random sample from a large population (Cf. Fixed Effect Model). A random effect model does not provide any knowledge of the treatment effect at a particular level but it does enable study of the variability due to the effect of treatment.

Random Sample

A simple random sample is a sample of individuals or observations drawn from a population where each has an equal chance of being selected.

290 Randomized Clinical Trial

Randomized Clinical Trial

Same as Randomized Controlled Trial.

### Randomized Controlled Trial

A clinical trial in which patients are allocated to treatments (including placebo) in a random fashion. The essential idea is that randomization removes confounding effects and reduces bias in the result. Also known as a randomized clinical trial.

Randomizing

Allocating patients to alternative treatments in a clinical trial in a random fashion (that is, by chance).

Range

The difference between the smallest and the largest in a set of numbers.

Rank Correlation Coefficient

See Spearman Rank Correlation Coefficient.

### Rank Dependence

A type of model that is often used in the analysis of decisions under uncertainty. In health economics it refers to a system for equity weighting quality-adjusted life-years (QALYs) in a social welfare function. The weight depends on each individual's relative ranking in terms of expected lifetime QALYs.

Ratchet Effect

An effect in negotiations between purchasers and providers by which the purchaser utilizes knowledge from previous negotiations to ratchet up the

Rationality 291

expected level of performance at the same price or global budget. See Purchaser-provider Split.

### Rate of Interest

Interest is the amount of money payable to a lender for lending a given amount for a period. The interest rate is that amount divided by the sum that is lent, usually on the assumption that the period is one year, and expressed as a percentage. See Discounting.

Rate of Return

The rate of discount (see Discounting) that makes the present value of a stream of money costs and benefits over time equal to zero.

Ratio Scale

A property of some measures of health and also of some measures of utility.

See Utility.

The idea that addiction may be explained in terms of the usual economic axioms of utility or expected utility theory. See Addiction.

Rational Drug Design

A focused strategy for organizing commercial pharmaceuticals research based on knowledge of the workings of proteins in human biology.

### Rationality

Generally intended in economics to refer to behaviour that is consistent with the axioms of utility or expected utility theory. This idea of rationality has been pretty well attacked. An early assault was from Keynes on the ground that its axioms were not obeyed by at least some financially successful people

292 Rationing who ought not to be dubbed 'irrational' in light of the evidence. It has been attacked also because it is too narrow in excluding emotional effects of not knowing things ahead of time, effects such as anxiety, disappointment and regret, and because it is too demanding and its literal pursuit might actually reduce welfare (see Bounded Rationality). Moreover identifying rationality as behaviour consistent with the axioms of a theory invites the unhelpful conclusion that every time any of these is actually violated (for example in a controlled experiment) the subject in question is an irrational being and, when this happens frequently - as it does - people are frequently irrational.

It is useful to keep in mind the distinction between positive and normative economics. In the former, rationality plays no ethical role at all, since theory seeks to account for or predict what actually happens, not what ought to happen if people were different from the analyst's postulates. Any empirical refutation of an axiom is a refutation of the theory (whether one calls it 'rationality' or anything else). In the case of normative economics, the issues are more comprehensive and complex than those that entail the usual notions of 'rationality' and certainly so if the concept requires people to be selfish and to act selfishly. In health policy, for example, 'rationality', like other fundamental value-laden underpinnings, needs to be defined and used in such a fashion as to include what the clients on behalf of whom the analysis is being done find ethically acceptable.

### Rationing

Allocating resources according to a rule or administrative arrangement. One rule might, for example, be 'resources shall go to whoever is willing to pay the highest price'. Such a rule does not much commend itself in health care however. The most common general usage of 'rationing' is in connection with (usually wartime) arrangements under which, in exchange for a voucher, individuals (or families) are entitled to purchase fixed quotas of goods at administered prices. A lot of tendentious hot air is generated in public debates about whether health care in any jurisdiction is 'rationed'. Those with political responsibility are understandably unwilling to concede that health care is rationed in either of the two ways just described, but sometimes less understandably unwilling to concede that some form of rationing mechanism has to be used, the critical question relating not to 'whether?' but to 'which?' There is also debate about the desirability of being explicit about the criteria to be used in determining the 'rules', with most economists apparently strongly in favour of explicitness. See Need, Pareto Optimality, Price Mechanism.

Rawlsian

An approach to questions of social justice named after the late American philosopher John Rawls (1921-2002). See Fairness.

RBRVS

Acronym for resource-based relative value scale.

Acronym for randomized controlled trial.

Acronym for risk difference. Real Income

Nominal (that is money) income adjusted to remove the effect of changes in the price level on purchasing power.

Real Price

Nominal price divided by a general price index. Recall Bias

A distortion in data that arises from people's imperfect memories of events they are being asked to remember. See Bias.

A plot of the sensitivity of a test against 1 minus the specificity of the test which is used to compare tests or to select an ideal cut-off value in a test.

294 Recurrent Cost

Recurrent Cost

Expenditures or opportunity costs that occur on a regular (usually annual) basis rather than being incurred once and for all.

### Redistributive Impact

The effect on the income distribution of the arrangements adopted in a jurisdiction for health care financing. It is generally measured as the difference between the Gini coefficients for prepayment and postpayment income distributions and, quantitatively speaking, seems to be determined by the average proportion of household income spent on health care, the progressivity of the health care financing system, the extent of horizontal inequity, and the extent to which households are reranked in the distributions when postpayment and prepayment distributions are compared.

### Reference Case

A standard set of methods and assumptions that analysts should follow in performing cost-effectiveness or cost-utility analyses. For two examples, see Marthe R. Gold, Joanna E. Siegel, Louise B. Russell and Milton C. Weinstein (eds) (1996), 'Appendix A: Summary Recommendations', in Cost-Effectiveness in Health and Medicine, New York and Oxford: Oxford University Press; chapter 5 in National Institute for Clinical Excellence (2004), Guide to the Methods of Technology Appraisal, London: NICE.

### Reference Cost

This pools data returned from (English) National Health Service (NHS) providers to compute average costs for all Healthcare Resource Groups and a Reference Cost Index for each NHS provider. It underpins the national tariff.

### Reference Cost Index

This is an index used in the English National Health Service (NHS) of the actual cost for each NHS provider organization divided by national average cost for the same activity multiplied by 100. An 'adjusted index' allows for local price variations.

Region of Acceptability 295

### Reference Pricing

A reimbursement mechanism (usually for pharmaceuticals) whereby a third party payer or insurer determines the maximum price at which it will reimburse the supplier (the reference price). The consumer pays the difference between the reference price and the market price. The reference price is often set at the price of the lowest priced product in the therapeutic group or is the weighted average of the lowest prices in the market. The market prices are set at the discretion of the supplier.

### Referral Bias

This occurs when particular physicians and centres of secondary care attract individuals with specific disorders or exposures but which are atypical of the general class of referrals or of patients of this type. Also known as centripetal bias. See Bias.

Referral Cue

Guidance to help General Practitioners and their patients decide when a consultation with a specialist, usually at a hospital, is appropriate.

### Reflection Effect

Reflection effects involve gambles whose outcomes are opposite in sign, though of the same size. Compare two choices, one between a certain gain of \$20 or a one-third chance of \$60 and the other between a certain loss of \$20 or a one-third chance of losing \$60. Most people choose the certain gain in the first choice but the one-third chance of loss in the second. The effect is predicted by prospect theory as a consequence of the S shape of the value function. Cf. Expected Utility Theory, Framing Effect.

Region of Acceptability

That part of a cost-effectiveness plane which indicates that the technology under investigation is cost-effective.

296 Regression Analysis