| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Research Reports |
CL Lewis, PT, PhD, is Assistant Professor, Department of Physical Therapy Education, Elon University, Campus Box 2085, Elon, NC 27244-2085 (USA) (lewisc{at}elon.edu).
M Moutoux, PT, MS, is Staff Physical Therapist, Home Care Providers of Alamance Regional Medical Center, Burlington, NC
M Slaughter, RPh, is Pharmacist, Home Care Providers of Alamance Regional Medical Center
SP Bailey, PT, PhD, is Assistant Professor, Department of Physical Therapy Education, Elon University
Address all correspondence to Dr Lewis
Submitted December 6, 2002;
Accepted July 11, 2003
 | Abstract |
|---|
Key Words: Falls • Home health agencies • Older adults
| Introduction |
|---|
 TopAbstract Introduction MethodResultsDiscussionConclusionReferences |
|---|
A "fall" is when a sudden, unintended loss of balance leaves the individual in contact with the floor or another surface such as a step or chair.1 Researchers1 have suggested that the increased incidence of falls in older people result form several factors. These factors include orthopedic, neurological, pharmaceutical, emotional and cognitive, and demographic variables. Orthopedic variables noted to increase the incidence of falls include: decreased force production by the hip muscles,4,5 decreased ankle range of motion,4 impaired mobility,1–3 and impaired gait (less motion and force, causing less toe-off and floor clearance),6 and decreased gait speed.7 Neurological factors such as slower reaction times,8 decreased visual acuity and visual perceptual skills,9,10 impairment in proprioceptive and vibratory sensory systems,11,12 impaired balance,8 and vestibular changes (dizziness, balance, and gait changes)13 are associated with an increased incidence of falls. Pharmaceutical-related variables are: taking more than 3 medications,14 orthostatic hypotension secondary to antihypertensive drugs,15 use of arrhythmic medications such as digoxin,16 use of psychotropic drugs,16 use of antidepressants and hypnotics,17 and use of tricyclic antidepressants.18 Researchers19 have noted that emotional and cognitive factors such as depression and impaired cognition seem to be related to increased incidence of falls. Researchers4,18,20 also have observed an increased incidence of falls in people aged 75 years or older. Although some researchers21,22 have reported a higher incidence of falls in women than in men, others23,24 found no differences in incidence of falls between men and women. Understanding factors that increase the risk of falls may assist in the identification of people who need interventions and allocation of resources to minimize the frequency and severity of falls.
The purpose of this exploratory study was to compare the characteristics of individuals who fell in 1999 while they were patients being served by a home health agency compared with a cohort of patients who did not report falling while receiving services from the same agency during the same time period. Based on prior research regarding falls in other populations,8,14–16,21 we theorized that patients who were older, of female sex, having more falls prior to admission to the home health agency, having comorbidities, taking more medications associated with increased risk of falling, and having fewer payer sources for health care were more likely to fall during the time of their home health service.
| Method |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
Admission Records
The records reviewed were the admission records for both groups and incident reports completed for all patients who reported that they fell while receiving home health care. At the time of data collection, the admission forms used were the Outcome and Assessment Information Set (OASIS) for individuals receiving Medicare. Those patients who did not receive Medicare had a Discipline-Specific Intake Report (DSIR) upon admission. Either nurses or physical therapists could make the initial visit and admit the patient for home health services. Although they contained similarities, the DSIRs were specific for each discipline.
The admission forms (OASIS or DSIR) for both the group with falls and the group without falls and incident reports of the group with falls were assigned code numbers. Only the staff of the home health agency retrieved the data from the record sources because of previously signed confidentiality statements. The patients' name, admission number, or incident report number was not recorded. Individuals falling more than once while receiving services were noted to have multiple falls.
Subjects
A total of 1,529 patients were admitted to the home health agency during the time period of the study (January 1–December 31, 1999). The people served by the home health agency resided within a county having a population of about 135,000 residents. Approximately 80,000 people lived within urban settings of small to moderate-sized towns, and the remaining 65,000 people lived in surrounding rural communities.
Of the 1,529 individuals admitted, 98 (6.4%) of the people fell during the course of their home health services for the study period. A group of 98 randomly selected subjects who received home health services during the same time period, but who did not report falling, was the comparison group. Random selection was achieved using the following method. A computer printout of the total number of admission records was made. The printout consisted of 16 pages (15 pages of 100 records on each page and 1 page with 28 records). Each page was assigned a number (1–16), and each record on the page was assigned a second number (1–100). Page numbers (1–16) were placed in one box, and record numbers (1–100) were placed in a second box. Random selection was achieved by pulling a page number and a record number. Each number was returned to the respective box before redrawing. If a page number and a record number drawn represented a record of a person in the group with falls, another set was drawn until a total of 98 records of people who did not fall were identified.
Variables
Variables chosen for analysis were those that were recorded on both types of admission forms for all patients. Variables investigated were identified as primary (referenced or supported) or secondary (exploratory and not previously reported). Primary variables included age,4,18,20 sex,21–24 primary and secondary diagnoses,4–13 and types of medications taken.14–18,25 Secondary variables included number of falls occurring during a 3-month period prior to admission to the home health agency, number of days of episode of care, number and type of disciplines providing patient care, and health care payer coverage. Of the possible variables documented to be associated with an increased risk of falling, these variables were chosen for the study because they were recorded and could be obtained for each subject. Although neuromuscular and musculoskeletal variables would have been of much interest, these variables were not routinely recorded for all patients. For the patients who fell while receiving home health services, we also investigated causes of falls and actions taken after a fall occurred.
For statistical purposes, the large numbers of diagnoses were categorized into 6 groups: neurological (eg, cerebrovascular accident, multiple sclerosis), cardiovascular (eg, congestive heart failure, peripheral vascular disease), respiratory (eg, emphysema, pneumonia), orthopedic (eg, total hip replacement, post-fracture), internal medicine (eg, diabetes, cancer), and functional limitations (eg, difficulty walking). At the time of our study, difficulty walking was a diagnosis on the OASIS and DSIF. These 6 categories were consistent with the International Classification of Diseases, 9th revision, 6th edition (ICD-9).26
The types and number of health care services were compared. The disciplines serving patients included skilled nursing, physical therapy, speech pathology, occupational therapy, and home health aides.
Medications investigated were benzodiazepines, antipsychotic phenothiazines, anticonvulsants, tricyclic antidepressants, narcotic pain relievers, and antihypertensives. These groups of medications were chosen because they are known to increase the risk of falls in subjects over the age of 60 years.25 Our determination of medications used did not reflect the total number or types of medications taken by each patient. In our study, we investigated only phenothiazines of the antipsychotic type, not all phenothiazines. Although subjects may have taken a number of drugs, only those falling into these 6 categories were considered in our analysis.
Subjects were placed into one of 4 payer groups for health care coverage. Subjects in category A had Medicare coverage only. Category B subjects had Medicaid only or Medicare supplemented with Medicaid coverage. Category C subjects had private insurance or Medicare supplemented by private insurance coverage. Subjects in category D were individuals who were self-pay or without payer coverage. The subjects in category D were not eligible for Medicare or Medicaid and were without private health care coverage.
Incident Report
Health care professionals of the home health agency completed an incident report when patients whom they were seeing reported that they fell. Patients may have fallen during an episode of care with a health care professional present or at other times when no health care professional was present. The health care professional seeing the patient at the time of the fall, or on the next visit if the fall occurred between a patient's visits, completed an incident report regarding the fall. The health care professional reported the nature and description of the fall, cause of the fall, and type of action taken after the fall. No distinctions were documented in incident reports completed for falls occurring during an episode of care or reported during a visit as having occurred previously. However, the health care professional did note in the report if he or she witnessed the fall.
Based on family or patient report of fall or direct observation of fall, the health care provider checked one of 4 designated categories listed on the report form for causes of falls: "1—not using assistive device or equipment correctly," "2—change in medical status," "3—safety issues (ie, lack of direct caregiver supervision at the time of the fall or environmental hazard in the home)," and "4—other (ie, unable to determine cause of the fall)."
In addition, the health care professional documented the action(s) taken after the fall occurred by checking one or more interventions from 9 designated categories listed on the incident report form. The categories were previously developed by the hospital personnel and not the staff of the home health agency or by us for the purpose of the study. The 9 categories included were: "1—patient education (ie, reinstruction in use of assistive device or equipment," "2—evaluation by physician," "3—additional services requested or additional intervention sessions," "4—notification of other disciplines and physician that the patient fell," "5—hospitalization," "6—patient placed in nursing home or other supervised living facility," "7—caregiver issues addressed," "8—pharmacy intervention (ie, medication re-evaluated)," and "9—other."
Data Analysis
Data were organized in Excel files.* All data were pooled and reported as group information for the group with falls or the group without falls. Unpaired t tests were performed on interval and ratio data to determine differences between the 2 groups with regard to age, the number of falls during a 3-month period prior to admission to home health services, the number of days home health services were received, and the number of disciplines represented by health care professionals seeing each patient. Chi-square analyses were used with categorical data to determine differences between the 2 groups with respect to sex, types of disciplines providing service, primary and secondary diagnostic categories, 6 categories of medications associated with increased risk of falling, and payer source for health care coverage. In the group with falls, frequency tables were used to characterize causes of falls occurring during home health services and actions taken after falls occurred. A Mann-Whitney U nonparametric statistic was calculated using only reported data in comparing the 2 groups. We used the Mann-Whitney U test because of the lack of a normal distribution for the data.
Means and standard deviations were calculated for categories. The level of statistical significance was set at P = .05. For variables that were different, the actual probability values are reported in the tables.
| Results |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
|
No differences were found for the primary diagnostic categories (Tab. 2), but there were differences in secondary diagnostic categories between the 2 groups (Tab. 3). For both groups, the primary diagnostic category was internal medicine followed by orthopedic disorders. The third most frequently documented primary categories were neurological disorders for the group with falls and respiratory disorders for the group without falls. For secondary diagnostic categories for both groups, internal medicine occurred most frequently followed by functional limitations. The group with falls, however, had an increased number of subjects with secondary diagnostic categories of neurological and cardiovascular disorders, whereas the group without falls had an increased rate of no secondary diagnoses (Tab. 3).
|
|
|
|
|
|
|
|
| Discussion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
We hypothesized that patients who were older or of female sex had more falls during the 3-month period prior to admission, had comorbidities, were taking more medications associated with increased risk of falling, and had fewer health care payer sources were more likely to fall.8,14–16,21 Data from the group without falls compared with data from the group with falls support the premise that older patients who were at greater risk of falling: (1) had more falls prior to admission, (2) had comorbidities of neurological and cardiovascular disorders, and (3) were taking more of the medications associated with risk of falling. The risk of falling, however, was not greater for female subjects or those having less health care coverage, or for different disciplines or lengths of episodes of service.
The fall rate we found, 6.4% of the total number of patients, was lower than that reported for people over the age of 65 years who are community dwelling.3,23,25,27–31 Several researchers3,23,27–31 reported fall rates of 25% to 35% in individuals over the age of 65 years. These fall rates, however, were determined over a period of 1 year or longer.23,28,29 The difference in fall rates we observed in our study may have resulted from the average time that the subjects were followed by the home health agency, which was approximately 2 months for both groups. Although the range for length of an episode of care was from 1 to 365 days, the average length of an episode of care was 60 days for subjects in both groups.
Almost twice as many female subjects as male subjects in both groups received home care services; however, the number of female subjects in the group with falls was almost identical to the number of female subjects in the group without falls. Thus, in our sample, the risk of falling was not sex-specific, which was similar to findings of other researchers.23,24,27 Campbell et al23 found no difference in fall rate between male and female subjects in their study of rural community-dwelling individuals over the age of 70 years. Several authors,3,21,28 however, have reported that women fall more frequently than men do. The reported differences in fall rate between sexes may have been related to the sample studied (ie, whether community dwelling or living in a nursing home facility) and the length of time the subjects were followed.
We did not find health care payer source to be associated with an increased risk of falling for our subjects. We analyzed health care coverage only as a factor for risk of falling. Medicare has age or disability qualifications. Medicaid has economic qualifications, and private insurance may suggest some degree of economic resource. In view of our data, however, we believe no inferences related to socioeconomic status and fall rate should be made based on health care payer sources. Socioeconomic data such as education level and income were not available in the patients' charts and thus could not be assessed in our study. However, the relationship of education, income, family resources, and social support to the risk of falling may be important to investigate in patients who are receiving home health care.
The characteristics that were different between the group with falls and the group without falls were the secondary diagnostic categories, medications associated with increased risk of falling, and number of falls during the 3-month period prior to admission. Each of these variables will be discussed individually.
Although no differences were observed related to the primary diagnostic categories, differences were observed in secondary diagnostic categories, suggesting to us that comorbidities may be a risk factor for increased rate of falls. The group with falls, however, had reported comorbidities of neurological and cardiovascular disorders more frequently than the group without falls. The group without falls had a higher rate of no secondary diagnosis. This finding suggests that comorbidities of neurological and cardiovascular disorders may be a factor influencing falls. People with a diagnosis of neurological disorders may experience more confusion, balance disturbances, muscle weakness, slower reaction times, and decreased endurance for ambulation that may increase fall rate.7,8,32 An increased risk of falling has been reported among patients with cerebrovascular accidents and neurological disorders.33,34 Various investigators29,32,35 have reported that when comorbidities were present, patients were at risk for less favorable health outcomes and increased risk of falling. People with cardiovascular comorbidities may experience generalized weakness or postural hypotension that may increase their risk of falling.15,16 Lawlor et al17 reported an increased risk of falls in patients who have an increasing number of simultaneously occurring chronic diseases, thus noting the effects of comorbidity and increased falls risk.
Our data indicate that subjects in the group with falls took more antipsychotic phenothiazines and tricyclic antidepressants than subjects in the group without falls. Examples of these medications taken by subjects in this study were Mellaril,
antipsychotic phenothiazine, and Elavil,
a tricyclic antidepressant. Tinetti and Ginter29 reported an increased rate of falling in patients taking antipsychotic phenothiazines and tricyclic antidepressants independent of other risk factors, including dementia and depression, the 2 disorders for which these categories of drugs are most commonly prescribed. Lawlor et al17 noted an increased rate of falling in patients taking antidepressants. In addition, patients who fell took a greater number of drugs in the 6 categories investigated.23,36
Fall occurrence during the 3-month period prior to admission was an indicator of a higher risk of falling. The group with falls had almost 3 times as many falls prior to admission as did the group without falls. Shumway-Cook et al37 investigated risk factors for falling and noted that a history of imbalance was one of the strongest indicators of future falls. From these data, it appears important for home care providers to gather information on falls that occurred prior to admission to the health care service and to note their frequency and alert other members of the health care team.
At the time of the data collection, the number of falls occurring during the 3-month time period prior to receiving home health care was a question on the OASIS and the DSIR. Because the current OASIS form no longer requires the reporting of falls that occurred prior to admission, this information is lost for future record reporting and patient identification.
When a fall occurred, more than 75% of the interventions for the subjects in our study involved patient re-education (review of patient transfers, gait, and safety instruction) (Tab. 8). Of the 98 subjects who fell, 20 subjects fell again after intervention, which accounted for 35% of the total number of falls. Thus, 20% of the time the actions taken after the fall were not successful in preventing a second fall. Possibly, the intervention of patient reinstruction was not adequate, subjects with additional medical diagnoses may have made errors in judgment regarding motor skills and motor planning ability, or the intervention may have reduced the number of falls but not eliminated all falls. People who fell may have attempted tasks beyond their skill level, or they may have had an inability to process information.
Incident report forms and data collected regarding falls for use in clinical management warrant additional research. Admission forms and incident report forms often dictate what information is collected. These forms, we contend, require close inspection for data collection because they could be rich resources of information or they could result in limitations in documentation.
Our findings were limited to retrospective data available for all subjects. Musculoskeletal factors such as the effects of patients' muscle force, balance, and gait speed on the risk of falls were not analyzed in this study because these data were not available for all subjects. Limitations of this study also include incomplete data entries on the OASIS. A prospective study could enhance documentation. We chose to limit the medication categories studied to medications known to increase the risk of falls and did not include all medication categories. Our study was limited by the sample size. Questions of the action of specific medications were not investigated because a much larger sample size would be needed to analyze each medication that patients were known to be taking.
For future research, prospective studies of people who fall while receiving home health services would be most important. Researchers could design the studies for documentation and data collection of specific musculoskeletal factors such as the effects of patients' muscle force, balance, and gait speed on the risk of falling. Although we noted no differences in fall rate among the different disciplines, we believe that in future studies it would be interesting to note separate intervention of nurses, occupational therapists, physical therapists, and speech pathologists on the risk of falling. The patient's psychosocial status and socioeconomic status regarding cognition, mental status, income, financial resources, education level, social support, and presence of family caregivers or other aid support are important characteristics that need to be investigated as possible factors in the risk of falling for this population.
| Conclusion |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
Falls remain a common and costly event for patients receiving services across the health care continuum. Development of profiles of patients receiving such services who are at greater risk of falling and development of successful intervention strategies to prevent falls are worthy of future study.
| Footnotes |
|---|
The project was reviewed and approved by the Committee for Human Subjects at Alamance Regional Medical Center and by the Institutional Review Board of Elon University.
* Microsoft Corp, One Microsoft Way, Redmond, WA 98052-6399. 
Novartis Pharmaceuticals Corp, 59 State Route 10, East Hanover, NJ 07936.
Astrazeneca Pharmaceuticals LP, 1800 Concord Pike, Wilmington, DE 19850-5437.
| References |
|---|
|
TopAbstractIntroductionMethodResultsDiscussionConclusionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
