Using patient sitters to directly observe patients at high risk for falls is a practice suggested as part of several evidence-based falls prevention guidelines. However, the clinical and cost-effectiveness of sitter programs has been questioned. Analysis of data from 75 hospitals participating in the Hospital and Healthsystem Association of Pennsylvania Hospital Engagement Network Falls Reduction and Prevention Collaboration revealed a statistically significant correlation (p < 0.05) between low rates of falls with harm and the use of sitter programs. A statistically significant correlation (p < 0.05) was also identified between low rates of falls with harm and three specific sitter program design elements: defining criteria for sitter qualifications, providing a training program for sitters, and establishing a pool of sitters. Analysis of falls reported to the Pennsylvania Patient Safety Authority by hospitals from across the commonwealth in which sitters were identified as being present at the time of the fall suggests that the use of sitters may be associated with a higher percentage of assisted falls and a lower rate of falls with harm. A review of the literature exploring the use of sitters as part of falls prevention programs, along with suggested strategies for sitter program design, is presented. A tool is provided to help hospitals make the business case for implementing and maintaining a sitter program.
Patient sitters (also called patient safety assistants, companions, and one-to-one or constant observers) are staff or volunteers assigned to provide direct observation of patients at risk to harm themselves or others. Patient sitters are used in a variety of care settings, including for patients who are assessed to be at high risk for a fall;1 in psychiatric crisis, including those who have attempted suicide or with suicidal ideation, or at risk for harming others;2 substance-abusing with behavioral problems;3 or experiencing delirium, confusion, or agitation.3
Patient sitters have been suggested in several evidence-based falls prevention guidelines4-7 and have become part of an arsenal of interventions used by healthcare institutions to provide close monitoring of patients at high risk to fall. The level of evidence supporting the inclusion of sitters in these evidence-based guidelines is expert opinion, which is generally considered less rigorous than evidence from systematic reviews and randomized controlled trials.8-10 Research into the clinical effectiveness of sitter programs has produced inconsistent results. In addition, the cost-effectiveness of these programs has been questioned due to the high costs associated with their maintenance.11
While the evidence base to support the use of patient sitters may be challenged, analysis of data from 75 Pennsylvania hospitals participating in the Hospital and Healthsystem Association of Pennsylvania Hospital Engagement Network (PA-HEN) Falls Reduction and Prevention Collaboration has found a statistically significant correlation between lower rates of falls with harm and the use of sitter programs, as well as specific sitter program design elements. Hospitals seeking to implement or sustain a patient sitter program that is both clinically effective and cost-effective are encouraged to incorporate specific sitter program design elements and consider existing strategies to structure sitter programs.
_______________* The analyses upon which this publication is based were in part funded and performed under contract number HHSM-500-2012-00022C, entitled “Hospital Engagement Contractor for Partnership for Patients Initiative.”
Patient Sitter Use and Falls-With-Harm Rates in Pennsylvania
Information about sitter program implementation and rates of falls with harm is available from 75 hospitals participating in the PA-HEN Falls Reduction and Prevention Collaboration. As part of the collaboration, hospitals completed the Hospital Engagement Network Falls Reduction and Prevention Collaboration Self-Assessment Tool (SAT) survey from July 5, 2012, through August 31, 2012. The falls SAT survey was designed to evaluate the current structure and content of hospital falls prevention programs compared with evidence-based, best-practice guidelines. The intent of the falls SAT survey was to assist hospitals in creating action plans targeted to the best-practice elements that were identified as missing or in need of improvement in their current falls prevention programs.
Hospitals completing the falls SAT survey were asked to report the level of implementation (i.e., no implementation, partial implementation, or full implementation) for individual falls prevention practices and falls prevention program elements across 17 categories of falls prevention practices. The use of patient sitters was the third lowest scoring category of practices. Forty-eight of the 75 hospitals reported having sitter programs, of which 21 reported full implementation of six specific sitter program design elements:
- A process for requesting and discontinuing sitters
- Patient eligibility criteria
- A pool of sitters
- Criteria for sitter qualifications
- A sitter job description with expectations for sitter behavior and responsibilities
- A training program for sitters
Rates of falls with harm were calculated for hospitals participating in the PA-HEN collaboration using the number of falls with harm (i.e., any fall requiring more than first-aid care12) as reported to the Pennsylvania Patient Safety Reporting System (PA-PSRS) and total facility patient-days as reported to the Pennsylvania Health Care Cost Containment Council (PHC4)13 for the period of January through June 2012. Analysis of falls SAT survey responses alongside rates of falls with harm revealed a statistically significant correlation (p < 0.05) between the use of sitter programs and lower rates of falls with harm. In addition, meta-regression analyses revealed statistically significant correlations (p < 0.05) between three specific practices and lower rates of falls with harm: the sitter program includes (1) criteria for sitter qualifications, (2) a training program for sitters, and (3) a pool of sitters.
The Pennsylvania Patient Safety Authority has previously published a full analysis of the falls SAT survey results identifying best practices associated with higher or lower rates of falls with harm. Although cause and effect cannot be established based on these analyses, the correlation between sitter programs and specific sitter program practices with lower rates of falls with harm were found to be statistically significant. In light of this correlation, as well as the low levels of full implementation for these practices reported on the falls SAT survey, the Authority has suggested that hospitals carefully consider sitter program implementation and incorporation of best practices in sitter program design.14
Sitter Use Associated With Increase in Assisted Falls and Decreased Harm
Falls are one of the events most frequently reported to the Authority.15 A query of the PA-PSRS database identified 54,289 falls events reported by hospitals across Pennsylvania from January 2012 through June 2013, of which 323 were identified as occurring with a sitter present. Assisted falls is one of the 13 PA-PSRS falls event types that hospitals are instructed to assign to falls that occur when a caregiver sees a patient about to fall and intervenes, lowering him or her to a bed or the floor. The percentage of falls labeled as assisted was much higher for falls occurring with a patient sitter present (54 of 323 falls, 16.72%) than for falls occurring without a sitter present (4,523 of 53,966 falls, 8.38%). Of note, none of the assisted falls with sitters present were reported as falls with harm (see Table 1). This suggests that using patient sitters may increase the chances of a fall being assisted, which in turn may prevent falls with harm from occurring.
Falls without Harm (% of |
Falls with |
Harm (% of
Total by Falls Event Type|
|Falls without sitter present||52,340 (96.99)||1,626 (3.01)||53,966|
| Unassisted || 47,853 (96.78) || 1,590 (3.22) || 49,443 |
| Assisted || 4,487 (99.20) || 36 (0.80) || 4,523 |
Falls with sitter present
|312 (96.59)||11 (3.41)||323|
| Unassisted || 258 (95.91) || 11 (4.09) || 269 |
| Assisted || 54 (100.00) || 0 (0.00) || 54 |
The following are examples of PA-PSRS reports of falls assisted by patient sitters:
Patient was sitting in a wheelchair with patient sitter in attendance. Sitter states patient was leaning forward in chair and sitter attempted to guard patient from fall, but patient proceeded to lean forward and was assisted onto floor.
Patient sitter stated the patient was sitting on toilet and started to lean to the left. [The sitter] grabbed the patient and eased him to the floor. The sitter called out for help, and the patient was found lying on the bathroom floor. His head was elevated off the floor.Sitter reported patient was sitting in chair and lost balance when patient went to get up. Sitter assisted patient to the floor.
Failure Modes Evident in Reports of Falls With Sitters Present
While sitter programs may reduce the likelihood of injury from falls, like many safety practices, they are not error-proof. Table 2 shows the percentage of falls occurring with sitters present as reported for each PA-PSRS falls event type. Assisted falls represented the most common falls event type assigned to these falls, but other falls event types (e.g., ambulating, toileting, found on floor) were reported at frequencies that suggest the need for clarification of sitter behavior responsibilities and expectations, as well as a sitter training program. Failure modes evident in falls reports involving sitters include the following:
The sitter was not within reach of the patient when the patient fell off a chair, wheelchair, or side of the bed.
The patient tripped on an item in the path of walking to the bathroom.
The sitter left the patient’s room with no designated backup staff, and the patient was later found on the floor.
The patient’s legs became weak while ambulating, and the sitter was in the patient room or in the hallway.
The patient was reaching for an item unassisted while sitting in a chair or wheelchair.
The patient was found on the floor after being left unattended in the bathroom while toileting or showering.
The patient slid to the floor while sitting on the edge of the bed.
Falls Event Type||
No. of Reports||
% of Reports|
|Found on floor||37||11.46|
|Lying in bed||36||11.15|
|Sitting in chair/wheelchair||26||8.05|
|Sitting at side of bed||20||6.19|
|Hallways of facility||7||2.17|
|In exam room/from exam table||3||0.93|
|* Total percentage does not equal 100 due to rounding.|||||
Information derived from the event reports suggests that a successful patient sitter program effective in reducing the number of falls that occur when a patient sitter is present includes specific sitter program design elements.
Sitter Program Design
Patient sitter guidelines vary among hospitals, including but not limited to guidelines regarding the duties of the patient sitter, when a patient sitter is requested, the method used to request a sitter, who can request a sitter, and when the sitter is discontinued. Hospitals may also differ in whether they allow the role of a patient sitter to be filled by clinical or nonclinical personnel, including nurses, nursing assistants, hospital staff, volunteers, or family members.1
For example, in a multihospital study by Torkelson and Dobal, a lack of collaboration and shared decision making was identified between caregivers and family members on when to begin, continue, and discontinue a patient sitter. The study concluded that sitter programs should be developed that (1) include policies and procedures clearly identifying the responsibilities of the staff nurses and the sitters, (2) provide clear instructions for the sitter, and (3) use psychiatric consults.2
Specific Sitter Program Design Elements
Facilities may be able to decrease patient sitter use while helping to reduce rates of falls with harm by incorporating the following specific patient sitter program design elements:
Designate staff responsible for overseeing the sitter program and/or assessing patients prior to initiating one-to-one observation (e.g., psychiatric liaison nurse, geriatric clinical nurse specialist).16-19
Outline a process for requesting and discontinuing sitters.2,8,16,20,21
Define patient eligibility criteria.8,16,20,22
Designate a pool of sitters.23
Outline criteria for sitter qualifications.21,24
Outline expectations for sitter behavior and responsibilities2,8,21,24,25 that include the following:
Reviewing pertinent clinical information, the reason for observation, and the plan of care with the nurse assigned to the patient and communicating regularly throughout the shift to report observed behaviors indicating either continued need for, or ability to discontinue, use of one-to-one observation2,23,25
Documenting observed behaviors and interventions provided in the course of performing one-to-one observation23,25
Maintaining toileting schedules for patients able to use the toilet or bedside commode5,19,25
Remaining with patients while in the bathroom4,25
Staying within arm’s reach of patients whenever appropriate (the nurse and the patient sitter will need to assess when remaining within close proximity to the patient may be inappropriate because it may increase agitation in some patients)25
Ensuring a safe environment (e.g., remove clutter, keep items within patient’s reach)4,5,6,7,19,25
Providing a proper handoff to another staff member, completed in the presence of the patient, when patient sitters must leave the patient25
Focusing on observation of the patient and avoiding non-work-related activities that may distract from care of the patient (e.g., personal calls, cell phone use, reading)25
Design a training program for sitters2,19,21,23,24 that provides education on the following:
Safe patient handling techniques4,7,25
Behavior management strategies for de-escalating agitated patients19
Diversional activities (e.g., activity aprons, crafts, magazine reading) to engage patients, particularly those with cognitive impairment19,23,25
Strategies to Structure Patient Sitter Programs and Control Costs
Patient sitters can be one of the costliest falls prevention methods, especially since the associated costs are not reimbursable from third-party payers.2 According to one national survey of 355 general medical-surgical hospitals conducted in 1996, annual expenses for patient sitters ranged from several thousand to a half-million dollars.26 In 2007, the Chester County Hospital, a 220-bed facility in West Chester, Pennsylvania, reported an unbudgeted cost of $515,480 in overtime and agency-hired nursing assistants attributable to the sitter program.23 These cost estimates may prove to be conservative, as more recent costs for sitter programs are not available in the literature.
Several strategies to structure patient sitter programs in order to reduce patient falls and control costs have proven to be effective for some healthcare facilities. Three of these strategies are outlined as follows:
Psychiatric liaison nurses. In an 800-bed Minnesota hospital, a lack of psychiatric resources for medical nurses who were not adequately trained in caring for patients with an underlying mental illness was identified as causing unnecessary use of patient sitters. In response, the hospital employed psychiatric liaison nurses (PLNs) who acted as consultants for medical patient care areas in the use of patient sitters.
PLNs facilitated day-to-day problem solving, provided education and support, and closely collaborated with the nursing staff on a management plan. A 50% reduction in sitter use was noted without increasing restraint prevalence, falls, or other adverse outcomes. The project also strongly suggested that a PLN can benefit medical nurses by “role modeling, teaching, encouraging, and supporting more effective and confident management of patients’ psychiatric needs.”27
A SAFE unit. Scripps Mercy Hospital, a 500-bed level I trauma center in San Diego, California, developed a “Specialized Adult-Focused Environment” (SAFE) unit. With the exception of acutely suicidal patients, patients who require frequent or constant observation are placed in this unit. Patients who do not require one-to-one attention and observation are placed in cohorted rooms next to or across from each other.
Newly hired employees are required to attend an eight-hour SAFE class and are called a “constant observer” or “therapeutic companion,” since their roles are viewed as more active than just sitting with a patient. After the first unit was established, other floors in the hospital established their own SAFE units. Falls rates were not reported, but a decline in sitter use and cost was noted.19
Patient sitter staff position. The Chester County Hospital created a new position of a patient safety assistant after realizing that their sitter expenses were increasing due to staff overtime and use of agency-hired nursing assistants. The average number of staff used to sit with the patients per month was 15.5 full-time equivalents (FTEs).23
At times, nursing assistants needed to be taken away from their assignments on patient care units to sit with a patient, causing the unit to be short-staffed, according to Angela R. Coladonato, MSN, RN, NEA-BC, senior vice president/chief nursing officer, the Chester County Hospital.23 “It caused huge dissatisfaction with the nursing staff,” Coladonato said in a telephone interview. “Their support person was now gone.”
She decided to “think outside the box” and formed a task force to identify ways to reduce the use of patient sitters. In the first phase, several initiatives were implemented, including the use of activity boxes, enactment of a new alcohol withdrawal order set with assessment guidelines, combining policies for constant observation of suicidal and nonsuicidal patients, instituting volunteer rounding, and establishing a mandatory sitter approval form initiated by nursing.
Coladonato worked closely with human resources and applicable managers, directors and educators to establish the patient safety assistant position. She started with 12 new FTEs and is currently budgeted for 16.23 FTEs. In order to minimize staffing costs, the position is designed to be part-time and does not include benefits.
When the patient safety assistant is not providing continuous observation, their duties may include restocking supplies for the department, delivering specimens to the lab, transporting patients, and responding to call bells. Table 3 outlines expectations for patient safety assistants compared with nursing assistants at the Chester County Hospital. (For more information, see “The Chester County Hospital Orientation Manual, Patient Safety Assistant, Nursing Staff Development,” available on the
Patient Safety Assistants||
|Documents on continuous observation form||X||Only when assigned to 1:1 observation|
|Handles patient positioning||X||X|
|Assists patients to bathroom||X||X|
|Changes bed linen for 1:1 patient (unoccupied bed)||X||X|
|Assists patient with meals as needed||X||X|
|Assists with bed-to-chair transfer (if applicable)||X||X|
|Transports patients ||Escorts patient when assigned to 1:1 observation or serves as a transporter when not assigned to 1:1 observation||X|
|Attends to general patient comfort needs||X||X|
|Engages patient with activity, if applicable||X||X|
|Reports patient behavior/movement to registered nurse||X||X|
|Assists clinical staff as requested for bathing/turning||X||X|
|Performs patient assessment (skin care/intake/output)||||X|
|Takes vital signs||||X|
|Provides incontinence care||||X|
|Restocks supplies for department||Only when not assigned to 1:1 observation||X|
|Delivers specimens to lab||Only when not assigned to 1:1 observation||X|
|Responds to call bells||Only when not assigned to 1:1 observation||X|
|Adapted with permission from the Chester County Hospital in West Chester, Pennsylvania. |
Falls-With-Harm Savings Calculator
The Authority has developed a
falls-with-harm calculator for hospitals that may be helpful in justifying the costs associated with implementing a sitter program, as well as other falls reduction initiatives. Hospitals enter their current rate of falls with harm, and the calculator provides the estimated average additional cost and length of stay associated with the current rate, along with estimates of cost savings that could be achieved with 10%, 25%, 50%, 75%, 90%, and 100% reductions in the rate of falls with harm.
The costs used in this calculation are operational costs incurred by hospitals both in treating injuries sustained in a fall and as a result of extended lengths of stay for these patients. These costs do not include additional costs related to malpractice claims or costs associated with subsequent outpatient care or readmissions.28 Such additional costs are not to be underestimated. A recent case was highlighted in the media29 that draws attention to the additional costs that may be incurred by hospitals that fail to provide sitter programs. The California Department of Public Health fined a hospital $100,000 in a case involving the death of a patient from intracranial hemorrhage following a fall because “there was no nursing supervision provided to the patient to ensure the patient was safe after the nursing staff had identified the patient had increased risk for falls and needed a sitter to prevent falls.”30
Falls identified as occurring with sitters present was determined through analysis of PA-PSRS report narratives. Of the 54,289 falls events reported by hospitals across Pennsylvania from January 2012 through June 2013, 666 reports included the term “sitter.” Of these, 323 report narratives described falls occurring with a sitter present at the time of the fall. It is therefore possible that this is an underestimate of the number of falls occurring with sitters present.
Data used in calculating rates of falls with harm is dependent on accuracy and consistency in reporting falls and identifying injury level within PA-PSRS. The 75 PA-HEN hospitals included in this analysis have agreed to a consensus definition for falls and falls with harm as a condition for participation in the PA-PSRS Falls Reporting Program; therefore, this limitation should be minimized. This data is also dependent on accurate and complete reporting of total facility patient days to PHC4.
Information on the implementation level for best practices in falls prevention was gathered from self-reporting hospitals completing the falls SAT survey. Designation of implementation level (i.e., no implementation, partial implementation, or full implementation) is subjective to the respondent.
Lastly, while meta-regression analyses have identified a statistically significant correlation between the use of sitter programs (and specific sitter program practices) and lower rates of falls with harm in 75 Pennsylvania hospitals, cause and effect cannot be inferred.
Patient sitters are one of many interventions that facilities can use to reduce falls. When implementing a patient sitter program, hospitals are encouraged to incorporate specific best practices in sitter program design. To structure patient sitter programs, hospitals can engage in strategies found to be successful in reducing falls and controlling costs in other organizations. The cost savings achieved in decreasing rates of falls with harm, both in terms of money saved and decreased severity of injury, may justify the costs associated with implementing and maintaining a sitter program.
Jonathan R. Treadwell, PhD, associate director, Evidence-based Practice Center, ECRI Institute, consulted on and contributed to statistical testing for this article.
- Morse JM.
Preventing patient falls. Thousand Oaks (CA): Sage Publications; 1996.
- Torkelson DJ, Dobal MT. Constant observation in medical-surgical settings: a multihospital study. Nurs Econ 1999 May-Jun;17(3):149-55.
- Shever LL, Titler MG, Mackin ML, et al.
Fall prevention practices in adult medical-surgical nursing units described by nurse managers. West J Nurs Res 2011 Apr;33(3):385-97.
- Degelau J, Belz M, Bungum L, et al.; Institute for Clinical Systems Improvement. Health care protocol: prevention of falls (acute care) [online]. 2012 Apr [cited 2012 May 15].
- Boushon B, Nielsen G, Quigley P, et al. Transforming care at the bedside how-to guide: reducing patient injuries from falls [online]. 2012 [cited 2013 May 8].
- National Center for Patient Safety. National Center for Patient Safety 2004 falls toolkit [online]. 2004 Jul [cited 2013 May 8].
- Agency for Healthcare Research and Quality. Preventing falls in hospitals: a toolkit for improving quality of care [online]. 2013 Jan [cited 2013 May 8].
- Capezuti E, Zwicker D, Mezey M, et al., eds.
Evidence-based geriatric nursing protocols for best practice. 3rd ed. New York: Springer Publishing Company; 2008.
- Greer N, Mosser G, Logan G, et al. A practical approach to evidence grading. Jt Com J Qual Improv 2000 Dec;26(12):700-12.
- Atkins D, Best D, Briss PA. Grading quality of evidence and strength of recommendations.
BMJ 2004 Jun 19;328(7454):1490.
- Harding AD. Observation assistants: sitter effectiveness and industry measures.
Nurs Econ 2010 Sep-Oct;28(5):330-6.
- Arnold TV, Barger DM. Falls rates improved in southeastern Pennsylvania: the impact of a regional initiative to standardize falls reporting. Pa Patient Saf Advis [online] 2012 Jun [cited 2013 Sep 30].
- Pennsylvania Health Care Cost Containment Council (PHC4) [website]. [cited 2013 Aug 5]. Harrisburg (PA): PHC4.
- Feil M. Falls prevention: Pennsylvania hospitals implementing best practices. Pa Patient Saf Advis [online] 2013 Dec [cited 2013 Dec 16].
- Pennsylvania Patient Safety Authority. 2012 annual report [online]. 2013 Apr 30 [cited 2013 Nov 5].
- Salamon U, Lennon M. Decreasing companion usage without negatively affecting patient outcomes: a performance improvement project.
Medsurg Nurs 2003 Aug;12(4):230-6.
- Worley L, Kunkel E, Gitlin D, et al. Constant observation practices in the general hospital setting: a national survey.
Psychosomatics 2000 Jul-Aug;41(4):301-10.
- Rausch D, Bjorklund P. Decreasing the costs of constant observation.
J Nurs Adm 2010 Feb;40(2):75-81.
- Nadler-Moodie M, Burnell, L, Fries J, et al. A S.A.F.E. alternative to sitters.
Nurs Manage 2009 Aug;40(8):43-50.
- Spiva L, Feiner T, Jones D, et al. An evaluation of a sitter reduction program intervention.
J Nurs Care Qual 2012 Oct-Dec;27(4):341-5.
- ECRI Institute. Do sitters make a difference?
Risk Manage Rep 2012 Feb;31(1):1-9.
- Tzeng HM, Yin CY, Grunawalt J. Effective assessment of use of sitters by nurses in inpatient care settings.
J Adv Nurs 2008 Oct;64(2):176-83.
- Coladonato A. Dealing with the patient sitter dilemma.
Bus Caring 2009 Feb;4(2):1-3.
- Lipkis-Orlando R, Mian P, Levy G, et al. Challenge for the 90s: a safe and cost-effective sitter program.
Medsurg Nurs 1993 Dec;2(6):483-5.
- Chester County Hospital and Health System.
Patient safety assistant orientation manual [hospital training materials]. West Chester (PA); Chester County Hospital and Health System; 2012 Dec:1-22.
- Worley LL, Kunkel EJ, Gitlin DF, et al. Constant observation practices in the general hospital setting: a national survey. Psychosomatics 2000 Jul-Aug;41(4):301-10.
- Rausch DL, Bjorklund P. Decreasing the costs of constant observation. J Nurs Adm 2010 Feb;40(2):75-81.
- Wong CA, Recktenwald AJ, Jones ML, et al. The cost of serious fall-related injuries at three Midwestern hospitals.
Jt Comm J Qual Patient Saf 2011 Feb;37(2):81-7.
- Clark C. Adverse events draw $775K in fines at 9 CA hospitals [online]. Health Lead Media 2013 Oct 28 [cited 2013 Nov 5].
- California Health and Human Services Agency Department of Public Health. [Statement of deficiencies and plan of correction for St. Jude Medical Center] [letter online]. 2010 Apr 2 [cited 2013 Nov 5].
Implications for Further Study
Hospitals participating in the Hospital and Healthsystem Association Hospital Engagement Network Falls Reduction and Prevention Collaboration completed the falls self-assessment tool survey for a second time in 2013. The results of the follow-up survey are being analyzed to assess whether or not the correlations found in 2012 persist between low rates of falls with harm and falls prevention practices, such as those for sitter programs.