Sunday, January 23, 2011

An informal telephone survey

Other than face-to-face discussion, what is the most effective way for me to communicate with my patients? Most frequently, I use the phone, but will occasionally send patients a letter, particularly if I’m reporting test results and recommending followup. I’ve dabbled with email, but found the “secure messaging system” available to me somewhat cumbersome.

Recently, I gathered data on how I was using phone calls to reach or respond to patients. Before I share the results and some interpretations, here’s how I collected the information.

From December 16, 2010 to January 14, 2011 (excluding my holiday week of December 20-24), I recorded data on all calls with patients and family members. I used the database program, Bento, on my iPhone. Bento makes it quick and simple to set up a basic data collection template (“library” in Bento-speak). Setting up a user-friendly template requires some forethought so as to make data collection simple. You can create a new library either directly on the iPhone, or on your computer and then sync the library between computer and iPhone.

The computer-version of Bento lets you search and sort the database, or export the information to a spreadsheet.

I prefer to limit the number of choices in each data field so that I can use pull-down lists and checkboxes. If all the information can be entered/confirmed with one touch on the screen, it speeds data entry and makes for consistent responses. I was interested in a snapshot rather than a lot of detail.

I used these categories:

Date – touch on date selects today’s date as default.

Duration of call (minutes) – I used the screen keyboard to enter the number. I could also have used a pull-down list of numbers from 1 to 20.

Reason for call – Pull-down list included:

Patient initiated – results (e.g. calling for lab reports)

Patient initiated – postop problem (e.g. concerned about infection after surgery)

Patient initiated – other question

Physician initiated – results (e.g. biopsy reports)

Physician initiated – followup (e.g. recovery after surgery)

Physician initiated – other

Time of call – one touch to record the current time


Call unsuccessful/busy/not in

Comments (always handy to have a “miscellaneous” column in a database!)

If you spend your whole work-day in front of your computer, you could just set up this database in Excel. Because I travel between several sites at several hospitals, in addition to my office, I’m much more likely to consistently record information on a single device, and my iPhone is always at hand.


Time period: 3 weeks

Number of calls: 39

Total call time: 179 minutes

Average time per call: 4.6 minutes (range 1-12 minutes)

Average call time per week: 59.7 minutes

Unsuccessful calls: 4 (10.3%) – (1,1,2 and 2 minutes respectively for these calls)

Average time per call (excluding unsuccessful calls): 4.9 minutes

Frequency distribution of successful call duration (chart below):

Local calls: 17 (43.6%)

Physician initiated: 16 (41%)

Time of day calls made (chart below):


My impression was that I made fewer calls than usual during the measurement period, likely because it was over the holidays. Less surgery is done during that time, so there would be fewer calls about post-op concerns. Also, it may be that patients were traveling, or busy with holiday visitors and so less likely to have time to call to discuss non-urgent concerns.

Concerns discussed in shorter calls (5 minutes or less) could probably be dealt with via email. These calls tend to be straightforward, single-question discussions. However, longer discussions are not conducive to email as there is a lot of back-and-forth with more involved discussion. Trying to conduct these discussions through email is probably not effective.

Three calls were quite lengthy (1 of 11 minutes and 2 of 12 minutes). Most of the time, I could anticipate that a call would be lengthy and so tried to make the call when I had adequate time for the discussion. However, this sometimes means that I would postpone making the call until I was going to be free for a longer period. I return many calls in the short breaks between OR cases or between seeing patients in the office. These short breaks aren’t adequate to discuss, say, a biopsy report showing cancer and the testing and treatment that will follow. These would be circumstances where scheduling time for a call (or office visit if convenient for the patient) would be helpful in reducing unsuccessful calls, and also making sure patients receive the information in a timely fashion.

In the study period, I spent about an hour per week speaking with patients on the phone. As I mentioned above, I think this was less than usual. This is one heck of a good deal for the patients’ insurance plan (AKA Sask Health) as I do not receive reimbursement for this work.

I’ve also been struck by the occasional patient’s comment that they are very pleased to be able to reach me by phone, as they didn’t think they could call directly to speak with “the doctor”. I know that many family practices insist that patients set up appointments to discuss concerns. If they didn’t have this rule, those family doctors could spend most of their day on the phone, gratis. But, if patients have been acclimatized to the idea that they can’t reach their doctor to ask questions, maybe some of them don’t even bother to make a call in the first place. If the problem were urgent, I suspect they would seek out help. But, for less-urgent problems, the questions may be going unasked. Perhaps I’m missing opportunities to provide better service/answer questions by not giving the level of accessibility that email would afford.