Objectives. To describe the attitudes of pediatricians and other physicians practicing in a fee-for-service environment toward compensation for telephone encounters with patients.
Design. Survey by mail.
Participants. The 67 pediatric and 634 other private physicians and medical school faculty of Albany County, NY.
Results. A total of 479 of all the physicians (68.3%) and 55 of the pediatricians (82.1%) returned the questionnaire. Of these, 69.9% of the total (95% confidence interval, 65.5% to 74.1%) and 58.2% of the pediatricians (95% confidence interval, 44.1% to 71.3%) indicated physicians should be compensated for calls with patients, especially for after-hours calls. This opinion was significantly associated with greater concern about liability for calls, more negative sentiments about after-hours calls, and a longer reported duration of calls. After adjusting for these factors, surgeons and pediatricians were significantly less likely to favor compensation than the group as a whole. Pediatricians in favor of compensation suggested charging a mean of $9.18 (SD $5.05) for 1 to 5 min, $14.00 (SD $8.87) for 6 to 10 min, and $22.27 (SD $12.62) for >10 min. Pediatricians reported documenting in patients’ charts a mean of 35.3% (SD 39.9%) of after-hours calls.
Conclusions. In a mostly noncapitated environment, the majority of pediatricians and other physicians favor compensation for telephone calls with patients. Some specialists, in particular pediatricians and surgeons, are, however, less likely to support this. Additional research into the reasons for these interspecialty differences may help to guide policy decisions on the financing of health care. telephone calls, physician compensation, physician opinion, health care system, survey.
Pediatricians and other physicians are seldom compensated for telephone calls with patients. Yet telephone encounters play an important role in medical care, especially when the office is closed. They constitute ∼20% of all encounters between patients (or their caregivers) and physicians.1 They provide patients and caregivers with important reassurance and advice,11,12prevent visits to the office or emergency room,8,13 and enable cost-effective follow-up of acute and chronic illnesses.16 Accordingly, it has recently been argued19,20 that paying physicians for telephone encounters, or counting them when measuring physicians’ productivity, would have multiple beneficial effects. It would give physicians fair compensation for the time, inconvenience, expertise, and liability risk involved in talking with patients by telephone. It could be expected to increase the efficiency of providing medical care and to improve the documentation of telephone discussions.
The health care system is, of course, rapidly evolving toward managed care, although payment of physicians by capitation is still far from the norm, especially for specialists.21 As such risk-sharing arrangements with physicians become more common, however, physicians’ incomes and the measures of their productivity will no longer be a function of the numbers of face-to-face encounters with patients, and they will lose the incentive to maximize such encounters. The telephone and eventually other modes of telecommunication are likely to be increasingly attractive ways to care for patients. Yet it will be many years, if at all, before most American physicians are paid by pure capitation. In the meantime, they will continue to work within a variety of fee-for-service arrangements in which telephone consultations, although a very common part of continuing medical care, are largely uncompensated.
What pediatricians and other physicians think and feel about telephone calls, whether they want to be compensated, and how well they claim they are documenting calls have been little studied. The purpose of this investigation was, therefore, to describe more fully the physicians’ point of view.
MATERIALS AND METHODS
The subjects of the survey were the physicians of Albany, NY, and its suburbs whose primary reimbursement from patients and third-party payers was through fee for service. These included both the private physicians and the full-time faculty members of the medical school. Although the latter are paid largely by salary, most of the revenues they contribute to the medical school faculty practice plan are compensations for individual services. Albany is a midsize city and the state capital of New York. The majority of its residents are enrolled in managed care plans that compensate physicians through fee for service rather than through capitation.
The survey was mailed to all subjects in early July 1994 along with a return envelope that was already stamped and addressed. A second survey and return envelope were again mailed to all subjects in late August. The questions are listed in the Appendix.
SAS was used to calculate descriptive statistics, simple correlations, and multivariate logistic regressions. Statistical significance was defined as P < .05. All reported confidence intervals (CIs) pertain also to the 95% level of significance.
Descriptive statistics were calculated for the entire sample, for each specialty, and for the subgroups of generalists and subspecialists within pediatrics, internal medicine, obstetrics–gynecology,and surgery. The proportions of physicians who favored compensation for phone calls as well as of responses to the other variables (representing physicians’ experiences with and attitudes about telephone consultations) were compared across specialties and subgroups.
Differences among physicians in their sentiments about after-hours calls were treated in two different ways. They were summarized by aggregating the five individual scores. In addition, however, because these aggregate scores were artificial constructs and because the individual scores were ordinal, rather than interval, and were variably linked to each other, the individual scores too were used in the comparative analyses.
Power calculations were performed by Instat software to examine the limitations resulting from the relatively small sizes of some of the specialties and subgroups.
Two multivariate models were then developed: a dichotomous choice (logistic) model to predict the probability of favoring compensation for telephone consultations and an ordinary least-squares model to predict the suggested dollar charges among those who wanted compensation. The models were used to infer what variables were independent predictors of physicians’ responses. The models were then examined to determine whether they adequately accounted for the observed between-specialty differences in the proportions favoring compensation and in the amount of compensation desired.
Tests of goodness of fit22 and of predictive accuracy (using the Receiver Operating Characteristic [ROC] curve23) were performed on the estimated logistic model to determine how well it described individual physicians’ opinions about compensation. The measure of goodness of fit compared the observed and expected proportions of “successes” (responses in favor of compensation) at each decile of estimated probability of favoring it. The ROC curve for the model was constructed by plotting the rates of true-positives and false-positives at different cutoff probabilities of favoring compensation. Because only 335 of the respondents answered all the relevant questions, this sample was used for the regression analyses; it did not differ significantly in demographic characteristics from the whole group of respondents.
Letters were sent to 701 physicians and returned by 479 (68.3%). The professional identities of the 459 respondents who indicated their medical specialties are compared in Table 1 with those of the entire surveyed population. Pediatricians made up a significantly higher proportion of the respondents than of the whole population (12.0% vs 9.6%).
The medical school faculty were represented equally among the surveyed population and respondents (29% of both). Both the population and the respondents were largely male, 84% and 82%, respectively. The mean age of respondents was 47.3 years (SD, 10.5); the median, 46; and the range, 31 to 81.
Opinions on Compensation
Only five physicians (1%) reported that they charged for phone calls; these included two internists, an obstetrician, an orthopedist, and a psychiatrist. Overall, 69.9% (95% CI, 65.5% to 74.1%) indicated that physicians should be compensated for calls with patients. More specifically, 63.0% of all the respondents indicated patient-initiated calls after hours, 43.2% patient-initiated calls to the office, and 18.4% physician-initiated follow-up calls.
Table 2 shows the responses to the general question about compensation according to specialty and groups of specialties. Internists and neurologists were significantly more likely than the combined sample of physicians to favor compensation, and surgeons were less likely. The lower than average percentages of pediatricians and general and family practitioners favoring compensation were not statistically significant because of the limited sample sizes. Gender, age, and faculty status had no significant effect.
Experience With Calls
Table 3 lists various types of telephone call experiences according to the respondents’ specialties. Variation among individuals was great, as indicated by the large SD values in the table. The overall reported length of phone calls averaged 5.7 min (SD, 4.1). The mean estimated numbers of phone calls were 6.5 per weekday night on call and 15.2 per weekend. General pediatricians reported that they spent the same time on each call (5.2 min) and covered the same number of physicians (3.6) as the aggregate of physicians but received significantly more calls when on call on both weekdays (mean, 14.4) and weekends (mean, 34.4). Documentation of after-hours calls is discussed further below.
Sentiments About After-Hours Calls
The physicians’ sentiments about after-hours phone calls were assessed on a 1- to 5-point scale in which a higher score indicated more negative feelings. Physicians found these calls moderately burdensome and annoying (medians of 3), although they were less inclined to feel abused by callers (median 2). Degree of annoyance was slightly negatively correlated with age (r = −.12, P = .02), but not with gender or faculty status. It was positively correlated with duration of call (r = .20, P = .0002) and number of calls on weekends (r = .19, P= .0006). Table 4 shows the ratings according to specialty. Pediatricians did not differ from the aggregate in spite of receiving the largest number of calls. Neurologists, however, appear to have been significantly more upset about after-hours calls than other physicians, and surgeons less upset.
Most physicians, however, felt that receiving after-hours calls, no matter how unpleasant, was part of their job (median 1) and did not make them “want to quit” (median 2). Nonetheless, a substantial minority (13.1% of the total and 31.3% of the general pediatricians) did report that these calls made them want to quit (ie, gave ratings of 4 or 5 on this item). These included 5 of the 7 faculty general pediatricians (71%) and 5 of the 20 private general pediatricians (20%) (relative risk, 3.6; 95% CI, 1.4 to 8.9).
Concern About Liability
Concern about liability for phone calls was moderate. On a scale of 1 = not at all to 5 = a great deal, the median degree of worry was 3 (interquartile range, 2). As shown in Table 4, there were no significant differences across specialties. Degree of worry was mildly negatively correlated with age (r = −.17, P = .0004) and with faculty status (r = −.13, P = .005). It was quite strongly associated with degree of annoyance with after-hours calls (r = .35, P = .0001).
Regression Analysis of Favoring Compensation
Multivariate logistic analysis identified three factors as independent predictors of physicians’ desire for compensation for telephone consultations: ratings of concern about liability for calls, aggregate ratings of sentiments about after-hours calls, and average length of calls (Table 5). An incremental increase on the 5-point rating scale for liability was associated with 1.55 (95% CI, 1.24 to 1.93) times greater odds of favoring compensation. In other words, a physician who rated his or her concern about liability at 5 had an odds of favoring compensation 5.76 (95% CI, 2.36 to 13.87) times as great as a physician who rated liability concern at only 1. Neither the number of calls received when on call (on weekdays or weekends) nor the number of physicians covered while on call was significantly associated with the probability of favoring compensation.
Both measures of statistical fit demonstrated that the estimated logistic model described quite well physicians’ opinions on compensation. The Hosmer–Lemeshow χ2 statistic computed for the test of goodness of fit was 4.36, with a correspondingP value of .82. The calculated area under the ROC curve was .81. The actual percentages of pediatricians and surgeons who favored compensation, however, were too low to be explained by the model (ie, they lay below the 95% CI for the predicated values). It is very likely, therefore, that factors other than those evaluated in this survey are important in explaining the differences among specialties in their opinions about compensation for telephone encounters.
Of the 198 respondents who wanted to charge for telephone consultations and also suggested dollar amounts to charge, the mean suggested charges, as shown in Table 6, were $10.86 for 1 to 5 min (SD, $6.73), $18.83 for 6 to 10 min (SD, $11.25), and $30.22 for >10 min (SD, $23.29). These correspond to ∼$2 to $2.50 per minute. The large SD values reflect the considerable differences among individual physicians. In general, pediatricians proposed lower compensations for telephone calls than physicians in higher-earning specialties, particularly surgeons, although the differences among specialties were not statistically significant.
A regression analysis showed that among the factors surveyed, concern about liability had the largest impact on suggested charges. Every 1-point increase in the rating on concern about liability was associated with a suggested charge that was higher by $1.33 (95% CI, $.53 to $2.12) for 1- to 5-min calls, by $2.16 ($.94 to $3.35) for 6- to 10-min calls, and by $2.55 ($.58 to $4.51) for calls of >10 min. Sentiments about after-hours calls had, after adjustment for the other variables, no impact on suggested charges. In addition, no correlation was found between amount of charges and a respondent’s age, sex, or faculty status.
Documentation of After-Hours Calls
As shown in Table 3, the percentage of after-hours phone calls that the physicians reportedly documented in the patients’ charts (based on the 432 who answered this question) was not high; the mean was 38.8% (SD, 38.2%) and the median 20% (interquartile range, 5% to 75%). The responses were widely spread; some physicians document well (24.8% estimated ≥75%), more document poorly (56.5% reported ≤25%).
A report of more complete documentation of after-hours calls was associated with a greater concern about liability for calls (P = .0005), holding other factors constant. Yet, even those who worried most about liability were likely to leave most calls undocumented; of those who rated their worry at 4 or 5, the mean reported rate of documentation was still only 45.5%. The percentage of documented calls was associated negatively with the number of calls received on weekdays (P = .03). Multivariate analysis revealed that after controlling for concern with liability and number of weekday calls, no significant between-specialty differences remained in the proportion of after-hours calls that pediatricians and other physicians said they documented in their patients’ charts.
The physicians’ descriptions of their on-call telephone experiences were in accord with the observations of other investigators. The reported average number of phone encounters when on call (Table 3) was 6.5 on weekdays and 15.2 on weekends, or an average of 1.7 and 4 calls per physician covered. General pediatricians reported the highest number of calls: 14.4 on weekdays and 34.4 on weekends (or 4 and 9.6 per physician covered). These rates are similar to those in other studies.11,14,24 The length of an average call (Table 3) was slightly longer than that found in the comprehensive Robert Wood Johnson study:12 5.7 min for the whole group (vs 4.6 min in the Robert Wood Johnson study) and 5.9 (vs 3.9) for pediatricians. Other investigators11,14,24have also reported shorter durations of calls than the estimates made by the survey respondents.
Most physicians in the Albany area stated that they should be compensated for at least some telephone calls with patients. The inverse may, however, be more surprising: 30% of all physicians and 42% of pediatricians did not think that they should be paid for telephone encounters, even though they know from experience that dealing with patients by telephone involves time, risk, and, often, inconvenience. Of those who did favor compensation, 89.6% affirmed this for patient calls after hours, 61.7% for patient calls during office hours, but only 26.3% for follow-up calls initiated by the physicians. They appear to see telephone access as part of the routine services offered to all patients who join a medical practice.28 Accordingly, even most of the respondents who favored compensation for after-hours calls felt that such encounters were “an expected part of my job.” Most pediatricians and other physicians may agree with the designers of Medicare’s resource-based relative value scale29 that follow-up calls are part of the service already charged for in the office visits.
Opinions about compensation differed among specialties. Internists and neurologists were more likely than physicians as a whole to want payment for telephone encounters (80.8% and 94.1%, respectively). Part of the variation can be accounted for by differences among specialties in those factors that were important independent predictors of whether individual physicians would favor compensation (such as worry about liability, sentiments about after-hours calls, and average duration of calls). In particular, neurologists were most upset by after-hours calls and reported the longest duration of calls. Yet, after adjusting for these factors, pediatricians and surgeons emerged as significantly less likely than physicians as a whole to want payment for telephone encounters.
This study provides no explanation of these residual differences among specialties. Nonetheless, it is easy to speculate about their sources. Surgeons may be accustomed to reimbursement for procedures rather than for talking. Moreover, many of the calls to them will concern these procedures (especially postoperatively), for which a single charge often includes the surrounding care; they are less likely than primary care physicians to be called by a patient about a new problem. Pediatricians, on the other hand, must give advice continually to parents on a multiplicity of important and unimportant topics and see this as part of their role. They may, in addition, be more sensitive than other physicians to the risks to their patients of any restrictions on access to care.
Making after-hours telephone care available to their patients does not necessarily mean that physicians must provide it themselves. Solo or small-group practitioners routinely share call with others. Furthermore, in Denver,26 82% of the pediatricians in private practice have subscribed, for at least some of their after-hours time periods, to an after-hours telephone service that manages their patients’ calls. The 18% who did not subscribe said that their patients’ parents “expected and needed” them to be available to provide advice after hours. Such a telephone service has been set up recently in Albany as well. Physicians who want their patients to have after-hours access but who also find answering after-hours calls a burden would be expected both to desire compensation for such calls and to be willing to pay to avoid the burden.
Almost all respondents indicated that no matter how unpleasant it was to receive after-hours calls, it was a part of their job. Nonetheless, some were, of course, more annoyed than others; 13% of the total and 31% of the general pediatricians reported that these calls made them want to cease practicing. Similarly, other investigators have reported that ∼25% of after-hours calls provoke hostile feelings.11,24 Although it is alarming that after-hours calls make so many general pediatricians think about quitting, this reaction was less prevalent among the community pediatricians (20%) than among the full-time faculty (71%). It is unknown how many pediatricians and other physicians actually retire from or change their practices to avoid after-hours calls.
Pediatricians as well as other physicians, especially those who receive a large volume of calls, admit that they do a poor job of documenting after-hours calls in patients’ charts. Similarly in Denver, 71% of pediatricians reported documenting “selected” calls in patients’ charts, but only 43% did “most” calls.26 This deficiency is understandable because of the inconvenience and expense involved in such documentation. Physicians are, however, constantly reminded and are surely well aware of the benefits of good documentation, namely to improve continuity of care and to decrease liability risks.30,31 Indeed, those who are highly concerned about their liability for phone calls do claim to document these calls more frequently than other physicians; nonetheless, even they report documenting less than half their calls.
The study has, of course, a variety of shortcomings. First, the sample may be unrepresentative of physicians as a whole. The 32% of the private physicians and medical school faculty who did not reply may differ in a systematic fashion from the responders. It is possible that those who want a change–ie, those who would like compensation for phone calls–would be more likely to send back the questionnaire. Furthermore, Albany area physicians may differ from physicians in other parts of the country and in more rural or more urban settings. Second, as already pointed out, the power to detect meaningful differences was low within the less common specialties, such as the neurologists and psychiatrists, and within the subgroups of generalists and specialists. Both of these shortcomings could be remedied by a much larger, multiarea survey, although at the risk of obtaining a smaller percentage of respondents among physicians who had no ties to the investigators.
Third, with a brief, written, anonymous questionnaire, it is impossible to ensure that the respondents understood the questions and intended their answers in the same ways. It is also impossible to do more than infer and speculate about why they answered in the ways they did. A more detailed and probing questionnaire would, however, have had a lower rate of response.
The findings of this study support the recent arguments19,20 for compensating physicians for telephone calls with patients. First, the large majority of pediatricians as well as other physicians would welcome such a policy. Second, telephone encounters have many of the characteristics of the office and hospital encounters for which physicians are routinely paid. When called by patients or their caretakers and even when making follow-up calls, physicians typically take histories, make tentative diagnoses, propose management plans, and offer advice. The significant predictors of favoring compensation for phone calls such as worry about liability, feelings about receiving calls, and duration of telephone encounters are equivalent to the stress and time dimensions of work that physicians rated when the resource-based relative value scale was constructed.32 Furthermore, when physicians suggested rates of compensation, they exhibited a conception of the monetary value of their time that was consistent with the current resource-based relative value scale for office visits. The suggested charges of ∼$2 to $2.5 per min are close for most physicians to their personal contribution to the charges for office visits (ie, after subtracting the staff’s time and the overhead). Members of higher-earning specialties proposed higher compensations for telephone calls, even if not fully equivalent to their higher incomes.
Third, compensating telephone encounters is likely to have a favorable impact on the nature, cost, and outcome of medical care. The low level of documentation of after-hours calls could be expected to rise if physicians had a financial incentive to do this. When physicians charge for encounters with patients, however they take place, they must be able to provide third-party payers with documentation to support their claims. Moreover, telephone encounters are efficient and inexpensive and, therefore, can save time and money for the patients and third-party payers whenever they are as effective as office or emergency room visits.17,18 Patients and their caregivers realize that telephone access to physicians benefits them, and, as physicians and their staffs are well aware, frequently want to substitute telephone for office care. In the context of a questionnaire about otitis media, $25 was the median amount that 148 parents bringing their children to a suburban office reported they would be willing to pay to avoid an office visit (Sorum PC, unpublished data, 1995). Furthermore, the policy might well be extended to physicians’ telephone consultations with other doctors. Even without the availability of full telemedicine capabilities, it is frequently more cost-effective for a primary care physician to talk with a specialist by telephone than to send the patient.
Like office visits, telephone encounters with patients could be compensated according to their complexity, not their duration. It would be simpler, however, to start with a payment schedule based on contact time. The physicians’ proposals of dollar amounts (Table 6) can serve as a basis for establishing a suitable schedule, with an extra charge to cover the administrative costs associated with billing for and keeping records of telephone calls. A payment schedule of $15 for 1 to 5 min, $25 for 6 to 10 min, and $35 for >10 min would seem not only fair but also acceptable to physicians. The expense should be borne largely or fully by third-party payers so that parents are not dissuaded from calling when they should.20
It appears that pediatricians, parents, and third-party payers should all welcome incentives to physicians to provide more acute and follow-up care over the telephone. Until capitation transforms how physicians’ income and productivity are determined, such incentives will be in the form of monetary and nonmonetary compensation for telephone encounters. Payments to physicians will be cost-effective, as long as any actual charges to patients and families do not limit telephone access unduly and as long as physicians and patients use the telephone judiciously.
Your specialty. Age. Sex. Are you a full-time faculty member at Albany Med?
Do you charge for telephone calls?
Should physicians be compensated for calls with patients? For what types? Check as many as you consider appropriate: patient-initiated, during office hours? patient-initiated, after office hours? physician-initiated, follow-up? other? if so, what?
How much should be charged for a call lasting 1 to 5 min? 6–10 min? >10 min?
How many minutes does an average phone call with a patient last?
When you are on call, how many physicians do you cover?
On average, how many phone calls do you receive from patients per week-day night on call? per weekend on call?
Rate on a scale of 1 to 5 your reactions to after-hours phone calls.
1 5 a. an expected part of my job not a part of my job b. not burdensome very burdensome c. not annoying very annoying d. I do not feel abused by callers I feel abused by these calls e. these calls do not make me feel like quitting they make me feel like quitting
What percentage of after-hours calls do you document in the chart?
Rate from 1 to 5 how much you worry about liability for phone calls, where 1 = not at all and 5 = a greatdeal.
We thank Karen Mourtzikos and Jeffrey J. Rufo for their invaluable assistance in data collection and tabulation, and Elizabeth Higgins, MD, and James Lambrinos, PhD, for their useful comments on an earlier version.
- Received June 24, 1996.
- Accepted August 12, 1996.
- Address correspondence to: Paul Sorum, MD, AMC Internal Medicine–Pediatrics, 724 Watervliet-Shaker Road, Latham, NY 12110.
↵‡ Dr Mallick is a senior economist at Pracon, Reston, Virginia.
- CIs =
- confidence intervals •
- ROC =
- Receiver Operating Characteristic curve
- ↵Mendenhall RC. Medical Practice in the United States. A Special Report of the Robert Wood Johnson Foundation. Princeton, NJ: Robert Wood Johnson Foundation; 1981
- National Center for Health Statistics. Physician Visits. Volume and Interval Since Last Visit, United States, 1980. Vital and Health Statistics, Series 10, No. 144. DHHS Publication No. (PHS)83-1572. Washington, DC: US Government Printing Office; 1983
- National Center for Health Statistics. Current Estimates from the National Health Interview Survey: United States, 1982. Vital and Health Statistics, Series 10, No. 150. DHHS Publication No. (PHS)85-1578. Washington, DC: US Government Printing Office; 1985
- National Center for Health Statistics. Physician Contacts by Sociodemographic and Health Characteristics, United States, 1982–83. Vital and Health Statistics, Series 10, No. 161. DHHS Publication No. (PHS)87-1589. Washington, DC: US Government Printing Office; 1987
- Roghmann KJ,
- Hoekelman RA,
- McInerny TK
- Evens S,
- Curtis P,
- Talbot A,
- Baer C,
- Smart A
- Wasson J,
- Gaudette C,
- Whaley F,
- Sauvigne A,
- Baribeau P,
- Welch HG
- ↵Hosmer D, Lemeshow S. Applied Logistic Regression. New York, NY: John Wiley and Sons; 1989
- Poole SR,
- Schmitt BD,
- Carruth T,
- Peterson-Smith A,
- Slusarsk M
- Fosarelli P,
- Schmitt B
- Copyright © 1997 American Academy of Pediatrics