CHI+MED logo banner

Designing the Safest Number Entry Interfaces

Key points

  • We have evaluated existing interfaces for entering numbers and shown that the style of interface influences the type of error committed and so the severity.
  • With existing designs, there is a trade-off between speed of entering a number and doing it safely.
  • Digit-based keypads are fast but likely to lead to errors that are out by 10 or more, whereas 5-key interfaces are slower but less likely to lead to large and so dangerous mistakes.
  • When keying errors are made, like trying to increase the number beyond the largest possible, they should be blocked with a warning that should be acknowledged.
  • It is easier to enter numbers without making mistakes if the interface is tailored, based on which numbers are most commonly used for the intended task.

There are many different designs of interfaces for entering numbers in use across many technologies from phones to medical devices. For example the interface might be like a telephone keypad with one key per digit. Alternatively, it might involve a series of arrow buttons used to move a cursor in a number and change the digits. Even the most common digit-based keypads can have different arrangements. For example the typical layout of digits on a telephone is different to that on a calculator. Not only are there different interface designs, there are also different input styles: numbers can be entered one digit after another, or by increasing a number on the screen until the target is reached. This difference also highlights the different ways that people can think about numbers: is it merely a string of digits, or a single whole thing that represents the value of something in the world? Number entry is therefore an interesting and complex task, with its own unique questions, distinct from text transcription.

Speed versus severity of errors
It is all too easy to accidentally enter a wrong number - who hasn't phoned a wrong number or got a silly result on a calculator at some point? To avoid patients being harmed due to mistakes in numbers being entered on medical devices, it is important that the user interface and style of input chosen minimises the frequency and severity of errors made.

We have evaluated several interface designs experimentally and shown that the style of interface does influence the type of error made as well as the severity of errors. For instance the numeric keypad, a type of interface where digits are entered one after another, offers fast number entry but keying slips that lead to missing decimal points or missing digits are easy to make with this kind of interface. Such errors are also often go undetected, unlike on some alternative interfaces, because with a numeric keypad, the interface does not make the person look at the screen — instead their eyes are more often on the keypad. These keypads frequently lead to errors where the final number is out by 10 or 100 or worse. Errors of such a magnitude are very likely to harm patients. Other interface styles are generally slower but lead to less severe errors when errors are made. For example, the so called 5-key interface (which has left and right arrows to move a cursor left and right over the number being formed, up and down arrows to change the digit at the cursor and an enter key) is slower but less likely to lead to large errors. When safety is more important than speed, it would be preferable.

Tailoring the design of an interface to the task
Based on the pattern of numbers used in infusion tasks in hospitals taken from the logs of devices, we have also shown how design guidelines can be determined to improve number entry interfaces for infusion tasks. Based on the data from one hospital, for example, four design guidelines might be:

  1. As 0 is the most common digit, entering a string of zeros should involve no more than one key-press per digit. WIth some interfaces, such as incremental interfaces, this may or may not be the case, depending on the state the device was left in.
  2. Entering the value 999 should be possible with, at most, three key presses, which again may not be the case with non-numeric keypads.
  3. Decimal points are used very infrequently, so entering a decimal point should require the user to request that functionality to prevent accidental slips.
  4. The numbers 1,000, 100, and 50 are used in nearly half of all infusions and should have dedicated buttons. This provides a good trade-off between having buttons for every common number that would then be hard to find, and having only digit buttons, which would be slower than necessary for the common values entered.

These heuristics are designed to ensure that the most frequent numbers are entered quickly whilst reducing the number of button presses. The latter may minimise the risk of key slips as with fewer key presses there is less opportunity for key slips. This may especially help with respect to decimal point errors.

We tested these ideas by adapting three existing interfaces to incorporate these guidelines, In laboratory conditions, the speed was either unaltered or increased by the changes. The number of key presses was reduced and error rates were not affected, suggesting that the speed increase is not due to people being less careful. The next step is to show that similar results are obtained in more realistic conditions.

Avoiding slips with numeric keypads
Many devices that require number entry just ignore the user pressing the decimal point more than once. If a person tries to correct the key presses [0 • • 7 5], by trying to delete a decimal point, they could turn it into 75 rather than 0•75. If the second decimal point was just silently ignored then the attempted correction would delete the only one registered instead. Instead all keys the user presses (including multiple decimal points) should appear on the display. Also entering numbers that violate the international guidelines for formatting numbers should be blocked with the user alerted. We found that doing this for decimal points can at least halve the rate of errors where a person enters a number that is ten or more times larger or smaller than that intended. Few medical devices do this, yet it would be easy to fix them — and normal (error free) use would be unaffected.

A related problem is due to devices only having space to display numbers with at most 8 digits, say. Applications often 'ignore' digit keys pressed after the display is full. Trying to correct the 9 digit number 123456786 to 123456789 with the sequence of keys [1 2 3 4 5 6 7 8 6 DEL 9] could delete the 8 if the last 6 had just been ignored. This would turn it into 12345679. Again the user should be alerted and the extra digit blocked.

Overall the important design lesson is that devices should always alert users when they have made a mistake, like entering too many digits or multiple decimal points, rather than silently ignoring the flawed key press. Otherwise the person may not notice their mistake, or not notice that it has been silently corrected and then mistakenly try to correct it.

Reducing the severity of mistakes
WIth 5-key interfaces, similar issues arise. Again user interfaces should block detectable errors rather than ignore them, and should give warnings that the user must acknowledge. The kinds of errors that should be blocked here include key presses that would increase a number above the maximum or decrease it below the minimum, as well as key presses that would move the cursor off the end of the display. We also found that key slips are less likely to lead to large number entry errors if the cursor starts at the left-most end.

A. Cauchi, A. Gimblett, H. Thimbleby, Simulation to Evaluate Alternative Approaches to Blocking Use Errors, Journal of Medical Devices, 6(1):017502, 2012. doi 10.1115/1.4026680

A. Cauchi, P. Oladimeji, G. Niezen, H. Thimbleby, Triangulating Empirical and Analytic Techniques for Improving Number Entry User Interfaces, EICS2014, 6th ACM SIGCHI Symposium on Engineering Interactive Computing Systems, 243–252, 2014. doi 10.1145/2607023.2607025,

A. Cauchi, P. Curzon, A. Gimblett, P. Masci, H. Thimbleby, Safer "5-key" Number Entry User Interfaces using Differential Formal Analysis, Proceedings BCS Conference on Human-Computer Interaction — BCS-HCI, XXVI:29–38, 2012.

A. Gimblett, H. Thimbleby, User Interface Model Discovery: Towards a Generic Approach, Proceedings ACM SIGCHI Symposium on Engineering Interactive Computing Systems — EICS 2010, G. Doherty, J. Nichols, M. D. Harrison eds., 145–154, 2010. doi 10.1145/1822018.1822041.

P. Cairns, H. Thimbleby, Reducing Number Entry Errors: Solving a Widespread, Serious Problem, Journal Royal Society Interface, 7(51):1429–1439, 2010. doi 10.1098/rsif.2010.0112.

Patrick Oladimeji, Harold Thimbleby, and Anna L. Cox. Number entry interfaces and their effects on error detection. In Proceedings of the 13th IFIP Conference on
Human-Computer Interaction. September 2011.

Patrick Oladimeji, Harold Thimbleby, and Anna L. Cox. A performance review of
number entry interfaces. In Proceedings of INTERACT 2013, IFIP Conference on
Human-Computer Interaction: Designing for Diversity. IFIP, September 2013.

Sarah Wiseman, Anna L. Cox, and Duncan P. Brumby. Designing devices with the
task in mind: Which numbers are really used in hospitals? Human Factors, 55(1):
61-74, 2013. doi:10.1177/0018720812471988. In top three for the HFES prize.

Sarah Wiseman, Orla Hennessy, Anna L. Cox, and Duncan P. Brumby. Tailoring
number entry interfaces to the task of programming medical infusion pumps. In
Proceedings of the Human Factors and Ergonomics Society (HFES) International
Annual Meeting, September 2013.