A descriptive study of managed-care hassles in 26 practices

Abstract

Objectives To explore the nature of managed-care hassles in primary care physicians' offices and to determine the feasibility of practice-based research methods to study the problem. Methods 16 internists and 10 family physicians volunteered to collect data about managed-care hassles during or shortly after the office visit for 15 consecutive patients using preprinted data cards. Outcome measures Number of hassles, time required for hassles, and interference with quality of care and doctor-patient relationship. Results Physicians adapted easily to using data cards. Before the pilot study, participants estimated a hassle rate of 10% and thought that interference with quality of care and the doctor-patient relationship was infrequent. Of 376 total visits for which the physicians completed data cards, 23% of visits generated 1 or more hassles. On average, a physician who saw 22 patients daily experienced 1 hassle lasting 10 minutes for every 4 to 5 patients. More than 40% of hassles were reported as interfering with quality of care, the doctor-patient relationship, or both. Conclusions The high hassle rate, in addition to the interference of hassles with quality of care and the doctor-patient relationship, suggests the need for further investigation into managed-care hassles using practice-based research methods.

As managed care becomes the dominant mode of health care delivery in the United States, physicians are expressing concerns about the effects of managed-care policies on the daily work of patient care.^1,2,3,4 Complaints focus on an array of issues that interject themselves directly into the doctor-patient visit. These include restricted formularies, limited access to medical specialists, the requirement of prior approvals for procedures, unavailable treatments, lengthy appeals processes, and physician payment delays. These intrusions, often presenting ethical dilemmas, may take valuable time away from the visit and undermine patients' trust in their physicians.^5,6,7,8 The phrase “managed-care hassles” has come to refer to these annoyances, which have been documented in “hassle hot-lines” and “hassle logs” and even reported to cause “hassle hypertension.”^{9,10,11,12,13}

In 1997 the Kaiser Family Foundation and the Harvard School of Public Health surveyed physicians and found that 87% had experienced some type of coverage denial for health plan services. Most thought that at least a third of denials resulted in potentially “serious” negative consequences for a patient's health.¹⁴ We found no studies that used patient visit-specific data, collected by the physician at the time of the patient encounter, to study the incidence of managed-care hassles. In the tradition of practice-based research, which seeks to identify and frame research questions relevant to primary care practice,^15,16 we undertook a prospective descriptive study of the hassle phenomenon.

METHODS

The impetus for the effort came from the creation of a practice-based research network for physicians in Catholic Healthcare West (CHW), an integrated health care system in the western United States. In 1996, through discussions at hospital medical staff and medical group meetings, clinical leaders in 3 California CHW regions and 1 Arizona region presented the concept of a practice-based research network and elicited topics for an initial network project. From among attendees, leaders selected 50 primary care physicians in 4 regions who they saw as busy, well-respected, open-minded clinicians. These physicians received a list of 10 clinical and practice management topics or questions generated at the initial meetings. They were asked to select and rank-order those topics that they would be willing to study in their offices to improve patient care and increase their satisfaction with practicing medicine. A third returned the survey with “incidence of managed-care hassles” ranked as first or second choice. “Diagnosing depression” and “patient understanding of medications” most frequently were the third and fourth choices.

The 50 physicians were invited to participate in the first network project by collecting data prospectively for 15 consecutive patients regarding the occurrence of managed-care hassles during an office visit. Based on literature review and our experience,^9,10,13 we defined a managed-care hassle as “any administrative matter relating to a managed-care organization's contractual arrangement between a patient and a physician that intrudes on or interrupts the patient visit.” Before starting data collection, the physicians completed a practice, sex, region, type of practice, and numeric estimates of the following indices: active patient volume, daily office visits, visit time, and managed-care volume. They were also asked to estimate how often they encountered a managed-care hassle; the time involved; the most frequent types of hassles; and how often hassles interfered with quality of care, the doctor-patient relationship, or both (see box for definitions).

Table 1.

Hassles that interfered with quality of care or the doctor-patient relationship
A hassle was defined as interfering with quality of care if:
A needed service was not available (eg, a specific drug was not on a health maintenance organization's formulary), or The effort required arranging the service reduced time the physician spent with the patient
A hassle was defined as interfering with the doctor-patient relationship if:
The physician's inability to provide a needed service had an adverse effect on the doctor-patient relationship, or The time and efforts required to arrange for the service had an adverse effect on the doctor-patient relationship

Of the 50 physicians originally sent surveys, 29 (58%) agreed to participate in the pilot. Three physicians were excluded because they had largely specialist practices. The resulting sample consisted of 26 physicians, a response rate of 52%, of whom 16 were internists and 10 were family physicians. The physicians practiced in 3 different CHW regions in California and Arizona. Table 1 describes physician practice demographics by specialty. Most physicians saw at least 22 patients per day (range, 10-30 patients) and spent 17 minutes or less per patient (range, 10-30 minutes). Most estimated that more than half of their patients had some form of managed-care insurance, but this varied considerably by region. (A patient had managed-care insurance if health care was provided by a contract between the primary care physician and a managed-care organization.)

Table 1.

Physician practice characteristics by specialty (N = 26)

	Practices, no. (%)
Characteristic practice	Internal medicine	Family
Number	16	10
Physician sex, male	14 (88)	9 (90)
Location of practice
Sacramento, CA	3 (19)	8 (80)
San Francisco, CA	6 (38)	0
Redwood City, CA	4 (25)	0
Santa Cruz, CA	2 (12)	1 (10)
Phoenix, AZ	1 (6)	1 (10)
Type of practice
Multispecialty	5 (31)	1 (10)
Single-specialty	9 (56)	7 (70)
Solo	2 (12)	2 (20)
Active no. of patients >600	12 (75)	3 (30)
Patients seen per day ≥22	12 (75)	4 (40)
Minutes spent per visit ≤17	13 (81)	8 (80)
Managed-care volume ≥50%	10 (62)	10 (100)

Between April and December 1997, physicians collected patient-specific visit data during or shortly after each visit using preprinted data cards. First, they recorded the patient's initials, age, sex, and whether the patient was new or a returning patient. Second, they indicated if they experienced a hassle during the visit. If no hassle occurred, data collection ended. If a hassle did occur, 4 additional data points were requested:

• hassle type: medication, laboratory test, imaging study, physician referral, nonphysician referral, or other;

• the physician's action(s) in response to the hassle;

• whether the hassle interfered with quality of care, the doctor-patient relationship, or both; and

• the approximate time taken to deal with the hassle during the visit.

Only 1 card per patient was completed, but more than 1 hassle could be described per card. To expedite data collection, physicians did not provide patients' insurance data. Physicians completed cards for 15 consecutive patients during 1 to 3 contiguous office sessions. The authors conducted brief telephone interviews with physicians during data collection to assess the ease of using the cards. On receipt of findings, the physicians were asked to complete an 8-question survey eliciting reactions.

The lack of systematic physician sampling precluded the use of statistical tests for significance to look for hassle rate differences as a function of practice factors (eg, physician specialty). Given the hypothesis-generating nature of the effort, we used descriptive statistics only—means, medians, modes, and ranges—to provide insight into the frequency and nature of managed-care hassles.

RESULTS

The physicians generated cards for 376 patient visits; 57% were with female patients, 35% with patients aged 65 years or older, and 4% with patients aged 14 years or younger. Fourteen percent of visits were with new patients. Physicians in Sacramento and San Francisco generated 65% of visits.

In telephone interviews with 23 of the 26 physicians, the physicians reported that cards were easy to use and required 60 to 90 seconds to complete; occasional problems included lost cards and delays in returning completed cards. The mean number of completed cards returned per physician was 14.5. Eighteen physicians returned 15 each; the remaining 8 physicians returned between 9 and 18. About 90% were completed within the originally suggested time-frame of 3 consecutive working days. Missing data were minimal, with fewer than 5 visits lacking data in any single category.

Table 4.

Selected comments from data cards (physician indicated that hassle interfered with quality of care and/or the patient-physician relationship)
Had to leave the room, get physical therapy referral list, found out the location nearest her house is not covered by insurance; will have to call to see if they will see her Patient had mammogram without authorization; time spent straightening it out Trying to refer patient to pool therapy program at hospital X; had to get clarification if health plan will cover the service Patient with 20-year history of bipolar disorder, no psychotropic drugs covered by plan Long-term patient, new to managed-care plan; medication that patient currently is taking is not effective in managing lipids, and medication that worked previously is not on formulary I get angry; patient gets anxious Patient upset because she may not be able to continue to see me under new HMO plan Referral was lost in paperwork; delay in repeating study for esophageal varices Patient wanted consultation with out-of-plan specialist at academic center X; had to tell patient that the consultation was not covered Needs home care visit; HMO didn't know who was contracting with them to provide home care this year; had to call around and inquire—delay in getting patient seen Patient changed plan; required new prescriptions written for all medications and also change in others due to new formulary; new drugs may not work as well as old drugs Patient with presumed ulcer disease—formulary restricts choice of medication; patient perception of “limited treatment” erodes doctor-patient relationship
HMO=health maintenance organization

Before the pilot study, most physicians estimated 1 to 2 hassles for every 15 visits, an expected hassle rate of 10%. They thought that hassles threw them off schedule 15 minutes or more every other day (as opposed to once or twice a day), that medication hassles were the most common, and that quality of care and/or the doctor-patient relationship were adversely affected “on occasion” (as opposed to “rarely,” “half the time,” “frequently,” or “almost always”).

For the 376 visits, the physicians reported a total of 97 hassles. The 97 hassles occurred during 86 individual patient visits, producing a point prevalence of 23%. Three physicians reported no hassle visits in their set of 15, and 7 physicians reported 5 or more. The median physician hassle rate for the 26 physicians was 20% (range, 0%-60%). Half of the physicians, therefore, reported at least 3 hassles for every 15 patients. For the 86 hassle visits, the mean time per hassle was 11 minutes (range, 2-38 minutes). The median hassle time per physician was 10 minutes (range, 8-14 minutes).

Physician referral problems were the most common hassles, representing more than a third of total hassles. From comments physicians wrote on cards, obtaining pre-authorization and coping with patient dissatisfaction when faced with restricted consultant choices were common referral problems. Medication problems represented a quarter of all hassles. Comments about medication hassles described confusion about coverage and frustration around denials for specific medications. At least 1 action was checked for each hassle. More than 1 action was checked in about 30% of hassle visits; these hassles took 15 minutes on average (table 2). Of the 86 hassle visits, 37 (43%) interfered with quality of care, the patient-physician relationship, or both. For 41 of the 86 visits with 1 or more hassles, physicians wrote comments on the card (see box).

Table 2.

Managed-care hassle characteristics and hassle times

Characteristic	Proportion of hassles, %	Mean hassle time, min
Hassle types (n = 97)
Physician referrals	35	10
Medication	26	5
Nonphysician referrals	11	5
Imaging studies	11	5
Laboratory tests	6	7
Miscellaneous*	11	10
Single-action hassles (n = 61)	71
After-visit work	46	14
Exit room	26	7
Phone call	12	4
Miscellaneous†	16	7
Multiple-action hassles (n = 25)	29	15
Hassle consequences (n = 86)
Interference, quality	30	13
Interference, relationship	16	5

^*For example, hospitalization delayed, procedure authorization needed.

^†For example, need to read medical record.

Half of the 26 physicians returned a post-pilot study questionnaire describing reactions to the findings. They acknowledged that knowing the hassle rate could help budget time, but most were not optimistic about solutions. Suggestions included switching physician groups, ending managed-care contracts, and becoming politically active “to regain physician control.”

DISCUSSION

Statement of principal findings

We quantified managed-care hassles for 26 primary care physicians; half the physicians reported a hassle rate of at least 20%. Most spent 10 minutes a hassle; those requiring multiple actions took as long as 15 minutes. Close to half the hassles had negative consequences for quality of care or the doctor-patient relationship.

Strengths and weaknesses of the study

The practice-based research methods used in this study were well accepted. Overall, physicians complied with the data collection protocol, used the data cards with ease, and found time to write comments on about half the cards. In this initial effort to launch a research network, at least 2 important requirements for practice-based research were set aside: identification of a large, representative physician pool of physicians from which to obtain a diverse, volunteer sample; and a formal enumeration of the physicians' individual practices as regards active patient volume and patient demographics, including health insurance type. Furthermore, not wishing to strain the capacities of office staff, data were collected during a single, short period of 1 to 3 days. These accommodations resulted in survey-based as opposed to actual estimates of active patient volume and proportion of managed-care patients. In addition, a small group of volunteer participants may inadvertently have biased patient selection and measurement of the hassle phenomenon during a single opportunity to collect data. (Investigator bias could be less of a concern because before the study began, participants expected a hassle rate that was half of what they actually reported.)

Future research

A follow-up study should address the pilot study's limitations by closely adhering to practice-based research requirements:

• identification of a diverse physician population from which to select a representative sample;

• appropriate methods for assuring high rates of physician enrollment and retention;

• systematic patient sampling for preventing selection bias;

• involvement of office staff, trained and reimbursed, to provide accurate denominator data for practice enumeration (eg, age, sex, or insurance registry), handle the “weekly return”¹⁷—an accounting of the total number of patients seen during the data collection period—and collect visit-specific data too time-consuming for physicians (eg, patient insurance); and

• development of sources other than physicians to verify hassle occurrence and effect on time, quality of care, and doctor-patient relationship (eg, patients could report hassles to determine whether physicians and their staffs buffer patients from plan restrictions' negative effects¹⁸).

The research design for a new effort should permit an in-depth examination of hassles as a function of practice characteristics, such as physician specialty and managed care volume. Internists and family physicians had similar hassle rates in the pilor (21% and 26%, respectively), but a possible threshold effect for managed-care volume was suggested. San Francisco and Santa Cruz physicians estimated their proportion of managed-care patients at about 40% and reported a combined hassle rate of 33% and a median hassle time of 12 minutes. Sacramento physicians, with 70% managed-care volume, reported a rate of 19% and a hassle time of 6 minutes. Are physicians less tolerant of hassles when they occur infrequently?

Meaning of the study: implications for clinicians or policymakers

With full appreciation of its limitations, to our knowledge, ours is the first report that has examined hassles using a prospective, visit-specific approach. Half of our participating physicians, on an average day of seeing about 22 patients, dealt with at least 4 hassles and added a minimum of 40 minutes to their workday. In the era of the 15-minute doctor visit, time spent on hassles can become a major irritant, particularly when quality of care or the doctor-patient relationship is adversely affected. The results of the pilot suggest the need for further work. We recommend a follow-up study to obtain an in-depth investigation of the frequency, cost, and consequences of managed-care hassles.