Instrument template · Sleep qualityUpdated September 28, 2026

PSQI eCRF template with the right field types

The Pittsburgh Sleep Quality Index mixes clock times, minutes, hours and rating scales. Getting the field types right on the eCRF is what makes the seven component scores computable, so this template uses time, number and single-choice fields exactly where the instrument needs them.

  • Time and number fields
  • 7 component scores
  • Global score 0 to 21

Free sandbox · No credit card · 21 CFR Part 11 aligned

Sleep quality · Month 1

During the past month:

Usual bedtime

23:30

Minutes to fall asleep

35min

Usual getting-up time

07:00

Hours of actual sleep per night

6.5h
Numeric answers, ready for component scoring

PSQI at a glance

  • Published by Buysse and colleagues at the University of Pittsburgh in 1989. It measures sleep quality over the past month.
  • 19 self-rated items (plus 5 questions for a bed partner or roommate that are not scored) combine into 7 component scores, each 0 to 3.
  • Components are summed into a global score of 0 to 21. A global score above 5 distinguishes poor sleepers from good sleepers.
  • Several items are clock times and durations, not rating scales, so field types matter.
  • The PSQI is copyrighted by the University of Pittsburgh. Check its terms for your use; commercial use generally needs a licence.

The instrument

How the PSQI turns answers into seven components

Unlike most questionnaires, the PSQI is not scored by adding up items. Its answers are first combined into seven clinically meaningful components: subjective sleep quality, sleep latency, sleep duration, habitual sleep efficiency, sleep disturbances, use of sleeping medication and daytime dysfunction. Each component is scored 0 (no difficulty) to 3 (severe difficulty), and the components are then summed.

Some components use a single answer; others combine several. Sleep latency, for example, combines the minutes taken to fall asleep with how often the participant could not get to sleep within 30 minutes. Habitual sleep efficiency divides hours of actual sleep by hours spent in bed, calculated from the usual bedtime and getting-up time.

Why field types matter

If bedtime is typed as free text ("around half eleven"), sleep efficiency cannot be computed and every record needs manual recoding. The template uses Capture's time question type for bedtime and getting-up time, number fields for minutes and hours with range checks, and single-choice fields for the frequency and rating questions. Everything the scoring algorithm needs arrives as structured data.

Capture also has a built-in calculator for minutes or hours between two date-times, which you can use if your protocol collects bedtime and wake time as full date-times (common in sleep diaries). For the PSQI's usual clock times, time in bed is derived in analysis along with the component scores.

PSQI global · Subject 006-0011

Global score

9 / 21

Poor sleeper (above 5)
0-5 Good sleeper6-21 Poor sleeper (above 5)
Sleep latency2
Sleep duration1
Habitual efficiency (81%)1
Daytime dysfunction2

Structure

The seven PSQI components

ComponentBuilt fromeCRF field types
1. Subjective sleep qualityOverall rating of sleep qualitySingle choice
2. Sleep latencyMinutes to fall asleep plus frequency of not sleeping within 30 minutesNumber (minutes), single choice
3. Sleep durationHours of actual sleepNumber (hours, one decimal)
4. Habitual sleep efficiencyHours asleep divided by hours in bedTime (bedtime, getting-up time), number
5. Sleep disturbancesFrequency of nine disturbance reasonsSingle choice per reason
6. Use of sleeping medicationFrequency of medication useSingle choice
7. Daytime dysfunctionTrouble staying awake and keeping up enthusiasmSingle choice

Structure only. Obtain the PSQI, scoring instructions and permission from the University of Pittsburgh.

Data quality

Range checks that catch impossible nights

A participant who reports 11 hours of sleep but 7 hours in bed, or 300 minutes to fall asleep, has usually mistyped. Range edit checks on minutes and hours raise an auto-query for the site to confirm with the participant, before the values reach a component score.

  • Range checks on minutes to fall asleep and hours of sleep.
  • Time fields for bedtime and getting-up time, not free text.
  • All scored items required.
Edit checks software
Vital signs · Systolic blood pressure

Edit checks / auto-queries

2
Auto-query

Type

Range High

Operator

Greater than

Value

180

Priority: High

Auto-query

Type

Range Low

Operator

Less than

Value

80

Priority: Normal

Query raised automatically

Value 192 violates limit (180). Please verify.

Scheduling

Monthly recall means monthly administration

Because the PSQI asks about the past month, give it no more often than monthly. Pair it with a nightly diary for week-by-week change, and a sleepiness scale for daytime impact.

Schedule of assessments · Sleep quality study
AssessmentBLM1M2M3M6
PSQI
Epworth (ESS)
Sleep diaryDDDDD

D = daily morning diary

Build a PSQI form with time fields and range checks

Test it with sample answers in the free sandbox before your ethics submission.

Build your PSQI form free

Data quality

Where PSQI data goes wrong, and how the form prevents it

On paper the PSQI is known for messy answers: ranges instead of numbers ("20 to 40 minutes"), times without AM or PM, and hours of sleep that exceed hours in bed. Each of those makes a component score uncomputable or wrong, and fixing them afterwards means asking participants to remember a month that has already passed.

Numbers, not ranges

Number fields accept one value, and the help text can ask participants to give their best estimate. That removes the "20 to 40" problem at source. Range checks then flag values that are implausible rather than impossible, so a site can confirm them with the participant.

Unambiguous clock times

Time fields store a 24-hour time, so an 11:30 bedtime cannot be confused with 11:30 in the morning. Participants who work nights or shifts need a note in the help text; some protocols exclude shift workers because the PSQI assumes a conventional night-time sleep period.

Consistency between sleep and time in bed

Hours of actual sleep should not exceed the time between usual bedtime and getting-up time. Check this during data review before lock, because a sleep efficiency above 100% will distort the component score and the global score with it.

Before go-live

PSQI setup checklist

Permission confirmed

Check the University of Pittsburgh's terms for your use and record licence details if one applies.

Field types correct

Time fields for clock times, number fields for minutes and hours, single choice for frequencies.

Range checks set

Minutes to fall asleep and hours of sleep checked for plausibility.

Bed partner questions decided

Keep or remove the five unscored questions to reduce burden.

Scoring algorithm documented

Component derivations written into the statistical analysis plan.

Monthly schedule

No more often than the one-month recall allows.

FAQ

Questions about this template

Something not covered here? Ask us directly.

How is the PSQI scored?

Answers are combined into seven component scores, each 0 to 3, which are summed into a global score of 0 to 21.

What PSQI score means poor sleep?

A global score above 5 indicates poor sleep quality.

How is sleep efficiency calculated?

Hours of actual sleep divided by hours in bed (from usual bedtime to getting-up time), times 100. It is then scored as a component from 0 to 3.

What recall period does the PSQI use?

The past month, so it should not be given more often than monthly.

Are the bed partner questions scored?

No. The five questions for a bed partner or roommate provide clinical information but are not part of the global score.

Is the PSQI free to use?

It is copyrighted by the University of Pittsburgh. Check its terms for your use; commercial use generally needs a licence.

Collect sleep quality data you can score

Time fields, number fields and range checks, ready in minutes. Free sandbox.

Build your PSQI form free