• Whether clients seek care,
• Where they go for care,
• Whether they are willing to pay for services,
• Whether people in need of family planning adopt a contraceptive method,
• Whether clients follow the provider's instructions on correct use,
• Whether clients continue using their method,
• Whether they return to the provider, and
• Whether they recommend services to others (78, 85, 94, 96, 116, 170, 183, 207, 367, 395, 399).

Health care clients often expect poor-quality care, accept it without complaint, and even express satisfaction when surveyed (52, 309, 316, 318, 319, 351). Client satisfaction, as expressed in interviews or surveys, does not necessarily mean that quality is good; it may mean that expectations are low. Clients may say they are satisfied because they want to please the interviewer (52, 267, 382), because they are afraid care might be withheld in the future (231), because of cultural norms against complaining (64), or because they respond positively to the word "satisfied" (309, 399).

Thus managers should not assume that the care provided is adequate just because clients do not complain. Even low levels of reported dissatisfaction—as low as 5%—should be taken seriously (382).

When clients' perceptions of quality are inaccurate, their expectations can influence providers' behavior and actually lower the quality of care (27, 116, 277, 282). Clients sometimes want inappropriate tests, procedures, or treatments in the mistaken belief that they constitute good quality (27, 116, 183). In response to such demands, Peruvian physicians have prescribed unnecessary medicines for childhood diarrhea (282), and Indonesian providers have given unnecessary injections to ill adults (28). In both cases providers were afraid that clients would switch to other providers if they did not receive the care they expected. Counseling clients and informing the public about what constitutes appropriate care are often important aspects of providing good-quality care.

Clients' satisfaction is an important indicator of service quality (379). Client satisfaction is difficult to assess, however (211, 309, 320). Data collection methods and measures often influence clients' responses (380). As noted, most clients say they are satisfied regardless of actual quality. Also, both access to services or health outcomes (for example, experiencing a side effect) can affect client satisfaction so much that it does not reflect the care-giving process (353).

To overcome these obstacles, researchers are exploring different ways to measure and analyze client satisfaction. Exit interviews can ask clients to report what happened during the consultation rather than to evaluate it (68, 381). In-depth interviews and focus-group discussions can probe for details rather than posing general questions about satisfaction (6, 14). Simulated, or mystery clients (that is, trained community members who pretend to seek services) can assess client satisfaction, on the arguable assumption that they share actual clients' perceptions (155, 156, 235, 319, 353).