The relationship between prices of services, quality of care, and patient time costs for general dental practice

Health Services Research, Feb, 1997 by Coralyn W. Whitney, Peter Milgrom, Douglas Conrad, Louis Fiset, David O'Hara

In contrast to medicine, dental care is still primarily delivered by community-based solo general practitioners. Few dentists are organized in larger practice units or are employed by health care organizations. Approximately 45 percent of dental care is paid for by commercial and nonprofit dental prepayment plans (Palmer 1995). Litfie dental care, except that provided under Medicaid, is paid for by the government. Dental fees have historically risen more rapidly than the consumer cost of living but less rapidly than the overall medical care cost index.

Nevertheless, as a result of the overall scrutiny of health care costs, pressure has been generated to control the growth of dental care prices. This is reflected both in greater consumer cost sharing and more restrictive benefits. In addition, insurers are increasingly seeking discounts from dentists included in Preferred Provider Organization (PPO) type arrangements. Many dentists feel strongly that the restrictions placed by insurers on covered services provided by the dentist and on dentist fees have reduced the quality of care provided to patients.

However, few studies have examined the issue of the basic relationship between price and quality in health services. Those available have focused on the technical quality of a particular procedure and the cost of that procedure (Weaver, Milgrom, Chapko, et al. 1985; Crall and Beazoglou 1989).

THEORETICAL STUDIES OF PRICE AND QUALITY DETERMINATION

DeVany and Saving (1983) showed that the economics of quality can be captured in three fundamental relationships: (1) the price facing the firm depends on the level of output and the set of product characteristics valued by consumers; (2) the firm's costs are also a function of the quantity and quality of its output; and (3) in turn, the firm's product quality depends on its quantity of output and capacity. The third relationship, which relates quality to the levels of actual output and potential output of the firm, defines quality as the reciprocal of the length of the queue for the firm's service. According to the theory, quality of service per unit of output may erode as the number of persons served or the quantity of service per person, or both, increases relative to the firm's fixed capacity. Alternatively, an increase in the firm's scale of potential capacity (e.g., space, number of operatories, and assistants) relative to the level of actual output would be expected to increase quality.

Klein and Leffler (1981) showed that in a world of repeat purchases of a given service, a price above the perfectly competitive level (equal to long-ran marginal and average cost) and the existence of firm-specific capital assets can act like a "guarantee" of superior performance by the firm. That is, the firm invests in assets such as advertising and reputation whose value is specific to the firm and can be realized only through a price premium. If the firm does not deliver service of sufficient quality to justify this price premium, its price will be lowered over time and the costs of specific investments will not be recovered. Thus, according to the theory, the price premium serves as a self-enforcing quality assurance mechanism.

EMPIRICAL STUDIES

Five previous empirical studies have addressed the issues of price and quality determination in health services. Each of these studies has proceeded from a theoretical model consistent with that of DeVany and Saving (1983). None, however, takes into account the potential role of price as a quality guarantee, as suggested by the theoretical framework of Klein and Leffler (1981).

In a study of demand for health care among the urban poor, Acton (1973) showed that the demand for care declines as patient waiting time increases. Holtmann and Olsen (1976) found a similar negative relationship between demand for dental services and waiting time.

House (1981) specified a "full price" model of dental care markets. He used ordinary least squares (OLS) regression to estimate a single equation with dental fee as the dependent variable and patient in-office waiting time, appointment delay, average dentist chairside time per patient visit, the dental assistant wage in the practice, dentist age, whether or not the state had reciprocal dentist licensure with other states, and the dentist-to-population ratio as independent variables. Waiting time and appointment delay were negatively related to the dentist's fee, and average dentist time per visit (posited to be a "quality" characteristic) was positively related to fee. The OLS model explained approximately 7 percent of the variance in dental fees.

One limitation of the model developed by House is that it treated waiting time, appointment delay, and dentist time per visit as fixed exogenous variables. In fact, these are influenced by the individual dentist's practice choices. Therefore, the estimates of the coefficients and standard errors by OLS in the price equation for these truly endogenous variables will be subject to simultaneity bias (Johnston 1972). Specifically, if a higher money price implicitly commits the dentist to delivering higher quality and lower patient time costs (more dentist time, lower office waiting time and appointment delay), then the waiting time and appointment delay coefficients will be biased and will not accurately reflect the true effect of those time costs on money price. Similarly, the coefficient of dentist time will be biased and will not accurately reflect the true effect of dentist time on the fee patients are willing to pay. These biases in estimation can be avoided only by structural models that take into account simultaneous causation. Second, the data source for House's model did not include quality of care measures, resulting in potential omitted variables bias in his estimates of the effect of time prices on fees.