The Patient Portal: What It Takes Off the Desk

A patient portal is almost always sold with the same promise: fewer phone calls. The promise is partly true. A portal reduces load significantly in certain areas, has almost no effect in others, and in one area even creates new work. It is worth knowing in advance what belongs where.

What really goes down

Appointment confirmation and reminders

This is the simple, worthwhile part. An automatic reminder with a confirmation button reduces outgoing calls and cuts no-shows. The effect on the calendar is also immediate: a patient who cancels three days in advance frees an appointment that can be filled.

Filling forms and questionnaires in advance

A health questionnaire, an intake form, a medication list — all of these are suited to being filled from home. The gain is twofold: less time in the waiting room, and above all information that arrives typed rather than in handwriting that has to be deciphered and re-keyed.

Documents and certificates

Requests for a visit certificate, a receipt, a summary — recurring requests that today go through the phone and email. When the patient can download them on their own, the request disappears from the queue.

Payments

Online payment of an open balance shortens collection and reduces reminder calls. Here too the effect is measured quickly.

What barely goes down

Medical questions do not disappear; they merely change channel. A patient who would have called to ask about a medication will write a message, and the message still requires a clinician to read and answer. This may be preferable, because the answer is documented and can be given at a convenient time, but there is no time saving here.

Complex appointment booking too — a treatment that requires two clinicians, a particular room or equipment — usually goes back to the front desk. A portal is excellent at standard appointments and mediocre at complex ones.

What creates new work

Releasing results. The moment a patient sees a result in the portal, they also see a value flagged as abnormal, and they call. This is not a portal problem but a policy problem, and it has three possible answers:

• Automatic release as soon as the result arrives. The most transparent, and the one that generates the most calls.
• Release after the clinician’s approval. Fewer calls, but it requires someone to review the results in time.
• Delayed release — for example after a number of days — unless the clinician released it earlier. A common compromise.

There is no single correct answer. There is an answer you must choose in advance, write down, and tell patients what to expect.

The rule: a portal shifts work, it does not erase it. It erases the work that is technical intermediation — delivering information that already exists — and leaves the work that requires professional judgment.

How to measure

Many clinics measure the percentage of patients who registered for the portal. That is a weak metric. Registration is not use. Four better metrics, all countable:

Metric	What it teaches
Percentage of questionnaires filled before arrival	How much data entry was saved at the front desk
Percentage of appointments confirmed digitally	How many outgoing calls were saved
No-show rate	The direct financial impact of the reminders
Average time to answer a portal message	Whether the new channel is managed or piling up

The last metric is the real test of success. A portal where messages wait a week drives patients back to the phone, and then you have two channels instead of one.

What about those who won’t use the portal

Even in a successful portal, some patients will not use it. Older people without a smartphone, those who do not read Hebrew, and those who simply prefer to speak with a person. This is not a failure but a planning fact. The practical conclusion is that you must not build a process that exists only in the portal. If the only questionnaire is digital, whoever did not fill it will arrive with no information and the staff will improvise. Better to keep a defined alternative path — a tablet at reception or a printed form — and key it into the same place.

A planning rule of thumb: if the portal usage rate is sixty percent, it means forty percent of patients go through the second path. The second path should be planned, not improvised.

Two things worth telling patients explicitly

• What the response time is. A single sentence on the messages screen saying how long a reply takes on working days prevents most of the frustration, even when the answer takes two days.
• What to do in an urgent situation. A clear instruction to call or to contact an emergency service, on every screen where a message can be written.

Cost versus benefit, in rough numbers

It is worth doing the calculation once. Count how many calls come into the clinic’s reception in a typical week, and how many of them belong to the four categories the portal handles: appointment confirmation, a document request, a balance question, a request to change details. Multiply by the average call duration. The number you get is the time you can get back, and not the promise in the brochure. For most clinics this comes to a few hours a week of front-desk staff, alongside a drop in no-shows — and that is usually enough to justify the move, even without grand promises.

Three decisions before launch

1. Who answers messages, and within how long. Written, not taken for granted. Including what happens during the vacation of whoever is responsible.
2. What is shown and what is not. Results, visit summaries, documents, a financial balance — each with a separate decision.
3. What the instruction is for an urgent case. The portal is not an emergency channel, and this must be stated clearly on every messages screen.

Adoption: where people actually register

The best registration happens at the front desk, at the end of the visit, when the secretary says one sentence and shows how to log in from the phone. An email campaign produces far lower percentages. It is also worth giving an immediate reason to log in: a questionnaire waiting, a document to download, a balance to pay. A portal with nothing to do on registration day will not be opened again.

And finally, accessibility. A significant share of patients will log in from a phone, some are older, and some read Arabic or Russian rather than Hebrew. A portal adapted to a small screen, with text large enough and in a language the patient understands, will get used. A portal built for a desktop screen will stay on the shelf, and the whole calculation above will not materialize.

Back to the blog