Online Appendix

Figure A-1. PubMed Search Strategy

("Sleep"[Mesh:noexp] OR "Sleep"[All Fields] OR "Sleep Deprivation"[Mesh] OR "Sleep Deprivation"[All Fields] OR "Sleep Deprivations"[All Fields] OR "REM Sleep Deprivation"[All Fields] OR "REM Sleep Deprivations"[All Fields] OR "Sleep Fragmentation"[All Fields] OR "Sleep Fragmentations"[All Fields] OR "Insufficient Sleep Syndrome"[All Fields] OR "Sleep Stages"[Mesh] OR "Sleep Stages"[All Fields] OR "Sleep Stage"[All Fields] OR "Drowsiness"[All Fields] OR "REM Sleep"[All Fields] OR "Sleep, REM"[All Fields] OR "Rhombencephalic Sleep"[All Fields] OR "Fast Wave Sleep"[All Fields] OR "Fast-wave Sleep"[All Fields] OR "Paradoxical Sleep"[All Fields] OR "Rapid Eye Movements"[All Fields] OR "Rapid Eye Movement"[All Fields] OR "Slow Wave Sleep"[All Fields] OR "Slow-wave Sleep"[All Fields] OR "sleep induction"[All Fields] OR "sleep time"[All Fields] OR "sleeping time"[All Fields] OR "Disorders of Excessive Somnolence"[Mesh] OR "disorders of excessive somnolence"[All Fields] OR "somnolence"[All Fields] OR "hypersomnolence"[All Fields] OR "hypersomnia"[All Fields] OR "hypersomnias"[All Fields] OR "diurnal sleepiness"[All Fields] OR "sleepiness"[All Fields] OR "Sleep Stages"[Mesh] OR "nonREM sleep"[All Fields] OR "non REM sleep"[All Fields] OR "daytime somnolence"[All Fields] OR "daytime sleepiness"[All Fields] OR "sleep latency"[All Fields] OR "sleeping"[All Fields] OR "night sleep"[All Fields] OR "nocturnal sleep"[All Fields] OR "rapid eye movement deprivation"[All Fields] OR "daytime sleepiness"[All Fields] OR "delta sleep"[All Fields] OR "stage 3 sleep"[All Fields] OR "slow sleep"[All Fields] OR "activated sleep"[All Fields] OR "deep sleep"[All Fields] OR "desynchronized sleep"[All Fields])

OR

("fast sleep"[All Fields] OR "low voltage sleep"[All Fields] OR "low voltage fast sleep"[All Fields] OR "low voltage fast sleep"[All Fields] OR "paradoxal sleep"[All Fields] OR "paradoxic sleep"[All Fields] OR "parasleep"[All Fields] OR "rapid eye movement sleep"[All Fields] OR "REM phase"[All Fields] OR "stage 5 sleep"[All Fields] OR "Sleep Initiation and Maintenance Disorders"[Mesh:NoExp] OR "Sleep Initiation and Maintenance Disorders"[All Fields] OR "sleep initiation and maintenance disorder"[All Fields] OR "Disorders of Initiating and Maintaining Sleep"[All Fields] OR "DIMS"[All Fields] OR "Early Awakening"[All Fields] OR "Nonorganic Insomnia"[All Fields] OR "Nonorganic insomnias"[All Fields] OR "Primary Insomnia"[All Fields] OR "primary insomnias"[All Fields] OR "Transient Insomnia"[All Fields] OR "transient insomnias"[All Fields] OR "Rebound Insomnia"[All Fields] OR "rebound insomnias"[All Fields] OR "Secondary Insomnia"[All Fields] OR "secondary insomnias"[All Fields] OR "sleep initiation dysfunctions"[All Fields] OR "Sleeplessness"[All Fields] OR "Insomnia Disorder"[All Fields] OR "Insomnia Disorders"[All Fields] OR "Insomnia"[All Fields] OR "Insomnias"[All Fields] OR "Chronic Insomnia"[All Fields] OR "chronic insomnias"[All Fields] OR "psychophysiological insomnia"[All Fields] OR "psychophysiological insomnias"[All Fields] OR "agrypnia"[All Fields] OR "agrypnias"[All Fields] OR "hyposomnia"[All Fields] OR "hyposomnias"[All Fields])

AND

("Hospitalization"[Mesh] OR "hospitalization"[All Fields] OR "hospitalizations"[All Fields] OR "hospitalized"[All Fields] OR "length of stay"[All Fields] OR "stay length"[All Fields] OR "stay lengths"[All Fields] OR "patient admission"[All Fields] OR "patient admissions"[All Fields] OR "voluntary admission"[All Fields] OR "voluntary admissions"[All Fields] OR "patient discharge"[All Fields] OR "patient discharges"[All Fields] OR "discharge planning"[All Fields] OR "discharge plannings"[All Fields] OR "patient readmission"[All Fields] OR "patient readmissions"[All Fields] OR "hospital readmissions"[All Fields] OR "hospital readmission"[All Fields] OR "readmission"[All Fields] OR "readmissions"[All Fields] OR "patient transfer"[All Fields] OR "patient transfers"[All Fields] OR "patient dumping"[All Fields] OR "hospital patient"[All Fields] OR "hospital patients"[All Fields] OR "Inpatients"[Mesh] OR "inpatient"[All Fields] OR "inpatients"[All Fields] OR "hospitalised patient"[All Fields] OR "hospitalised patients"[All Fields])

OR

(“hospitalized patient”[All Fields] OR “hospitalized patients”[All Fields] OR “in-hospital patient”[All Fields] OR “in-hospital patients”[All Fields] OR “in-patient”[All Fields] OR “aged hospital patient”[All Fields] OR “aged hospital patients”[All Fields] OR “aged hospitalized”[All Fields] OR “hospital admission”[All Fields] OR “hospital admissions”[All Fields] OR “hospital admittance”[All Fields] OR “hospital admittances”[All Fields] OR “hospital discharge”[All Fields] OR “hospital discharges”[All Fields] OR “patient transport”[All Fields] OR “intrahospital transfer”[All Fields] OR “hospital stay”[All Fields] OR “short stay hospitalization”[All Fields] OR “short stay hospitalizations”[All Fields] OR "night care"[Mesh] OR "night care"[All Fields])

AND

(“Adult"[Mesh] OR "adult"[All Fields] OR "adults"[All Fields] OR "aged"[All Fields[ OR "elderly"[All Fields] OR "aged, 80 and over"[All Fields] OR "oldest old"[All Fields] OR "nonagenarian”[All Fields] OR "nonageniarnans"[All Fields] OR "octogenarian"[All Fields] OR "octogenarians"[All Fields] OR "centenarian"[All Fields] OR "centernarians"[All Fields] OR "frail elderly"[All Fields] OR "frail elder"[All Fields] OR "frail elders"[All Fields] OR "functionally-impaired elderly"[All Fields] OR "frail older adult"[All Fields] OR "frail older adults"[All Fields] OR "middle aged"[All Fields] OR "middle age"[All Fields] OR "young adult"[All Fields] OR "young adults"[All Fields] OR “aged hospital patient”[All Fields] OR “aged hospital patients”[All Fields] OR “aged hospitalized”[All Fields] OR “very elderly”[All Fields] OR “very old”[All Fields] OR “aged patient”[All Fields] OR “aged patients”[All Fields] OR “aged people”[All Fields] OR “elderly patient”[All Fields] OR “elderly patients”[All Fields] OR “elderly people”[All Fields] OR “senior citizen”[All Fields] OR “senior citizens”[All Fields] OR “senium”[All Fields])

Figure A-2: Risk of Bias Assessment Tool for Non-Randomized Studies

Three non-randomized study designs:

  1. Pre intervention-post intervention
  2. Non randomized trial with a control group
  3. Non randomized with a historical control

General areas and specific questions to assess risk of bias

(adapted from the Suggested risk of bias criteria for EPOC reviews1 and the AHRQ Methods Guide for Effectiveness and Comparative Effectiveness Reviews2)

  1. Representativeness of the entire group
  2. Is it representative of a general inpatient population? yes / no
  3. excludes nursing home or ICU populations
  4. includes medicine, surgery, psychiatry, other gen inpt.
  5. Selection of the comparison group

a)are the groups comparable at baseline? (rate “no” as Major vs minor differences) yes / no (Mjr) / no (mnr)

  • If no, was there statistical adjustment for baseline characteristics?
  • Minimal demographics to include are:
  • age, gender, type of illness (e.g. pneumonia, chf, uti, etc)
  • baseline level of sleep disturbance (if they do a baseline before intervntn; some can't)
  • think about whether the environment could have been different

b)If using a historical control,

  1. Is it drawn from the same community as the exposed cohort? yes / no
  2. Is the historical control close in time and seasonality to the intervention group? yes / no
  3. Is there any reason to suspect that there could be differences making them not appropriately comparable? yes / no
  • If yes/ yes/ no then get a point here, otherwise not
  1. comparability of cohorts (confounders, effect modifiers; exclude from pre-post studies)
  2. Should control/adjust or account for the following important parameters that could affect outcome of sleep time/quality:
  3. drug or alcohol withdrawal
  4. advanced dementia, delirium, altered mental status
  5. obstructive sleep apnea
  6. sleep medications (ambien, benzodiazepine, benadryl, trazodone, haloperidol or other antipsychotic)
  7. Does the study address at least all of the above? yes / no
  8. Blinded outcome assessment (the people collecting data can't know the allocation, or should be blinded to the nature of intervention, or otherwise prevented from being able to figure out what group the patient is in) ?

a)Was blinded outcome assessment conducted? yes / no

b)If no, was there an objective measurement (actigraphy/PSG)? yes / no

c)If no: for self - report sleep questionnaires: were study personnel in the room or could have otherwise influenced the reporting (if they were in the room as the patient filled out the questionnaire, they could have influenced participant responses). yes / no

  • If no/ no/ yes: no points. They need to be able to answer yes to at least one of the first two or no to the last to get a point here.
  1. Completeness of follow-up

a)Have they accounted for all the patients that enrolled in the study? yes / no

  • They need to say what happened to the people that aren’t in the analysis

b)How big was the loss to follow up? What %?

  • If large, that can introduce bias.
  • If >10% = fair bias, > 20%=high bias 10% / 11-20% / > 20%
  1. Is there an adequate description of the intervention and how it was delivered? yes / no
  2. Reporting bias

a)were potential outcomes prespecified by researchers? yes / no / ?

  1. Are there other sources of bias that raise concern for the validity of the findings? yes / no
  1. Suggested risk of bias criteria for EPOC reviews, Cochrane Efective Practice and Organization of Care Group. Available at . Accessed September 14, 2013.
  2. Methods Guide for Effectiveness and Comparative Effectiveness Reviews. Agency for Healthcare Research and Quality. Available at Accessed September 14, 2013.

Qualitative Algorithm for Assessing Risk of Bias in Non-Randomized Studies

Figure A-3. Algorithm for rating the strength of evidence

Table A-1. Grade of trials assessing the efficacy of relaxation techniques on sleep quality compared with a usual care control among inpatient populations

Number of Trials; Subjects / Domains Pertaining to Strength of Evidence / Magnitude of Effect and Strength of Evidence
Risk of Bias / Consistency / Directness / Precision
RCTs and nonRCTs / Low SOE of an effect
8; 508 / Medium / Consistent / Direct / Imprecise / 0% to +38% improvement in sleep quality

SOE = Strength of Evidence

Table A-2. Grade of trials assessing the efficacy of reduced sleep interruption and sleep hygiene programs on sleep quality compared with a usual care control among inpatient populations

Number of Trials; Subjects / Domains Pertaining to Strength of Evidence / Magnitude of Effect and Strength of Evidence
Risk of Bias / Consistency / Directness / Precision
Non-RCTs / Insufficient SOE of an effect
2; 588 / High / Consistent / Direct / Imprecise / +5% improvement in sleep quality or quantity

SOE = Strength of Evidence

Table A-3. Grade of trials assessing the efficacy of bright light therapy on sleep quality compared with a usual care control among inpatient populations

Number of Trials; Subjects / Domains Pertaining to Strength of Evidence / Magnitude of Effect and Strength of Evidence
Risk of Bias / Consistency / Directness / Precision
Non-RCTs / Low SOE of an effect
3; 58 / High / Consistent / Direct / Precise / +7% to +18% improvement in sleep quantity

SOE = Strength of Evidence