ORIGINAL STUDIES
Time and Economic Cost of
Constipation Care in Nursing Homes
Lori Frank, PhD, Jordana Schmier, MA, Leah Kleinman, DrPH, Reshmi Siddique, PhD, MSc, Cornelia Beck, RN, PhD,
John Schnelle, PhD, and Margaret Rothman, PhD
Objective: Chronic constipation is a common disorder
among residents in long-term care; yet the cost to the
nursing home of constipation-related care is not
known. The objective of this study was to quantify the
nursing staff and supply-related cost of constipation
care to nursing homes from the perspective of the
nursing home.
Design: Prospective, observational time-and-motion
design.
Setting: Two United States nursing homes.
Participants: A total of 59 nursing home residents with
chronic constipation and nursing staff providing con-
stipation care to them.
Measurements: Actual time to complete constipation
care-related tasks was measured via stopwatch by
trained observers, and the number and professional
level (eg, staff nurse, CNA) of staff performing each
task was recorded. Frequency of constipation care task
data was obtained through 60-day retrospective med-
Constipation is a widespread disorder, and prevalence in-
creases with age. Between 30% and 40% of people aged 65 or
over report constipation problems,1,2 and up to half of all
nursing home residents suffer from the condition.3,4 Available
Research scientist (L.F., L.K.) and project manager (J.S.), MEDTAP Interna-
tional, Inc., Bethesda, Maryland; assistant director, Janssen Pharmaceutica,
Titusville, New Jersey (R.S.); professor, Department of Geriatrics, College of
Medicine, University of Arkansas for Medical Sciences, Little Rock, Arkansas
(C.B.); professor, Anna and Harry Borun Center for Gerontological Research,
University of California, Los Angeles, California (J.S.); executive director,
Health Economics, Johnson & Johnson, Titusville, New Jersey (M.R.).
Financial support for this research was provided by the Janssen Research
Foundation.
Address correspondence to Lori Frank, PhD, MEDTAP International, Inc., 7101
Wisconsin Avenue, Bethesda, MD 20814. E-mail: Frank@MEDTAP.com.
Copyright ©2002 American Medical Directors Association
DOI: 10.1097/01.JAM.0000019536.75245.86
ORIGINAL STUDIES
ical chart review for each subject. Nurse wage rate
data was obtained from the Nursing Home Salary and
Benefits Report, a US-based national source. Resident
and nursing home descriptive information was also
collected.
Results: The average cost per task occurrence ranged
from $0.72 for enema administration to $1.74 for oral
medication administration. Average nursing staff
costs per subject per year were $1577 for oral medica-
tion administration, $215 for dietary supplement ad-
ministration, $39 for constipation assessment, $17 for
suppository administration, and $6 for enema admin-
istration. Based on estimates of frequency of occur-
rence, the total annual labor and supply cost per long-
term care resident with constipation was $2253.
Conclusion: Nursing staff performance of constipation
care-related tasks is time consuming and costly in the
long-term care setting. (J Am Med Dir Assoc 2002; 3: 215–
223)
Keywords: Constipation; costs; staff burden
research estimates vary largely due to differences in definitions
of constipation, limiting estimation of the true magnitude of
the problem.
There are multiple causes of constipation, some of which
are particularly relevant to institutionalized older adults.
Among these are immobility, medications, endocrine or met-
abolic problems, neurological disorders, clinical depression,
eating disorders, structural anomalies of the colon or rectum,
inadequate fiber intake, and repressed or ignored urge to
defecate due to inconvenience or incapacity.2,3,5,6 Constipa-
tion may lead to additional morbidity such as hemorrhoids,
anal fissures, rectal prolapse, fecal impaction, and fecal incon-
tinence, each of which may reduce patient and caregiver
quality of life.3,7,8 The disorder as well as the sequelae carry a
cost in terms of staff time to diagnose, treat, and manage.
Constipation care tasks are largely the responsibility of
nursing staff in long-term care; yet there is limited informa-
Frank et al. 215
tion in the literature about nursing staff time required for care
of constipation. Care can include treatment for the disorder as
well as preventive care measures. Results of a small descriptive
study demonstrate that nursing staff find constipation care
moderately burdensome and time consuming and that some
constipation care tasks, such as disimpaction and enema ad-
ministration, can take close to one-half hour of staff time.9
Despite the prevalence of the disorder among institutionalized
older adults, no studies directly address the costs associated
with the disorder in nursing homes.10
The goal of this study was to quantify the nursing staff-
related cost of constipation care to nursing homes from the
perspective of the nursing home. An observational approach
was chosen for this study based on the success of the method
for a similar disorder, incontinence, within the long-term care
literature. Information about the length of time needed to
complete constipation-related tasks, the frequency with
which these tasks are completed, material resources used in
constipation care, and different responsibility levels of the
staff involved (eg, charge nurse, certified nursing assistants
(CNAs)) were gathered through direct observation and med-
ical record review. These data were used to obtain an estimate
of the costs associated with constipation care in the long-term
care setting.
METHODS
This was a prospective, observational, time-and-motion
study. Data on nursing staff time required for constipation-
related resident care was obtained via observation, along with
staff level involved (eg, charge nurses, staff nurses, and
CNAs). Observational constipation care data for residents
who met eligibility criteria were supplemented by medical
record data to estimate frequency of constipation care tasks
per resident.
Sites
Sites were included only if they met the definition of
“nursing home” as used in the Medical Expenditure Panel
Survey Nursing Home Component:11 A facility or distinct
portion of a facility certified as a Medicare skilled nursing facility;
or as a Medicaid nursing facility, or licensed as a nursing home with
three or more beds that provides onsite supervision by RN, or LPN,
24 hours a day, 7 days a week. The director of nursing (DON)
at each site supplied site descriptive information including
bed size, occupancy rate, and staff wage rates.
Study Sample
A sample of residents with chronic constipation were cho-
sen as the level of analysis rather than a sample of nursing staff
to maximize capture of constipation care tasks. Study staff
reviewed resident medical records and spoke to nursing per-
sonnel to identify residents who met inclusion and exclusion
criteria. Specifically, at Site 1 we went through records for all
patients residing in the facility at the time of our study and
selected for recruitment all patients who met eligibility re-
quirements. Because full record review of all patients was not
possible given the size of Site 2, we instead identified all
patients on constipation medications or otherwise eligible on
216 Frank et al.
several units within the site. Subjects were required to reside
in the nursing home for at least 60 days before enrollment,
and meet criteria for constipation, specifically, two or fewer
bowel movements per week for the prior 2 weeks (adapted
from the Rome criteria),12 and/or use of a stimulant or os-
motic agent, and/or enema, and/or bulk-forming laxative at
least two times in prior 4 weeks. Data were collected from a
convenience sample of two separate nursing homes in differ-
ent geographic regions (Mid-Atlantic and Northeast) in
1999. Institutional Review Board (IRB) approval was ob-
tained from a central IRB and a local IRB as required by one
site. Informed consent for participation was obtained from all
subjects or their guardians.
Data on demographic and clinical characteristics were ab-
stracted from the medical record. The following subject de-
scriptive data were collected: age, gender, race, bowel move-
ment frequency and body mass index (BMI). Any
hospitalization that occurred during the 60-day review period
was noted along with hospital length of stay. Use of site-
specific bowel treatment program was also recorded for each
subject.
Information on level of cognitive impairment was calcu-
lated based on Minimum Data Set (MDS)13,14,15 items in-
cluded in medical record review. A composite MDS cognitive
impairment score was calculated using dichotomous scoring
(0 ⫽ intact, 1 ⫽ impaired) for each of the nine MDS items
(short term memory, long term memory; recall of season,
room location, staff names/faces; presence in nursing home;
easily distracted; altered perception; and mental function vari-
ability). Mini-Mental State Examination (MMSE) data were
collected if available, and then only if recent (in prior quar-
terly assessment). Level of dependence for activities of daily
living (ADLs) was collected from Section G of the MDS, and
was calculated based on the method presented by Castle et
al.16 Possible ratings for each ADL are: 0, independent; 1,
supervision; 2, limited assistance; 3, extensive assistance; and
4, total dependence. ADL score was averaged across all 5
items (bathing, feeding, dressing, transferring, and toileting)
and reported as a dichotomous variable, with scores between
below 3 categorized as “low” and scores of 3 and above
categorized as “high.”
Both sites had internal protocols for care related to bowel
functioning, and whether the subject was formally on such a
protocol was also recorded. The protocols indicated what ther-
apeutic steps to take to maintain good bowel function and
specified care to institute should bowel movement frequency
drop below 1 per 3 days.
Data Collection
Observational data collection and medical record abstrac-
tion were performed by trained staff from a survey research
organization. All three observational data collectors were
registered nurses (RNs) with clinical and research experience
in inpatient settings. Observers also abstracted medical
records along with an additional non-RN abstractor.
Nine constipation-related care tasks (both preventive care
and treatment) were included on the data collection form,
with an “other” category available if other constipation re-
JAMDA – July/August 2002
lated tasks were observed or if multiple tasks were performed
simultaneously. Those nine tasks were (1) assessment of con-
stipation (review of bowel functioning, physical examination,
rectal examination, medical history from residents, consulta-
tion); (2) preparation and administration of special dietary
supplements for constipation; (3) toileting (bowel movement
only); (4) fecal disimpaction; (5) enema administration; (6)
suppository administration; (7) oral medication administra-
tion; (8) evaluating for response to treatment for constipation;
and (9) tasks that result from fecal incontinence such as
changing bed linens, bedpads, or clothes. Explicit written
indicators of task initiation and termination criteria were
provided to all observers to ensure consistent interpretation of
tasks. Observers and medical record abstractors were in-
structed to minimize disruptions to nursing staff at the sites.
The bowel protocols at each facility included monitoring
tasks (Task 1) and use of dietary and medication interven-
tions (Tasks 2, 6, and 7).
Task Observation
The goal of the task observation was to determine nursing
staff time and nursing personnel involved in completion of
constipation care tasks. The observers collected data on each
subject a total of four times, for 3 hours per time, a total of 12
hours of observation per subject. The schedule was determined
based on review of similar time and motion or cost studies
conducted on urinary incontinence (UI)8,17–20 and based on
expected frequency of target care tasks. Observation included
only hours between 7:00 AM and 7:00 PM, based on staff input
from the sites as to likely times for constipation care.
To help desensitize staff and residents at each site to the
presence of study staff, a desensitization period of 2 to 3 days
was used, during which study staff collected data that were not
included in the study.
A maximum of seven subjects were observed by a single
observer during each 3-hour observation bin. Observers could
watch more than one resident at a time because constipation
care tasks were neither continuous for each resident nor
overlapping in time for observed subjects.
For each separate constipation task observed, the following
information was recorded: staff time and level of staffing
required for each care task type (eg, staff nurse, CNA), aver-
age time for each care task (via stopwatch), and supplies
required for each care task. Information on the types of
supplies and medications (including type, amount, and fol-
low-up doses for medications) was also collected for compre-
hensive calculation of average task costs to supplement the
task labor cost calculation.
Medical Record Abstraction
The goal of the medical record review was to determine the
frequency with which the specified constipation care tasks
were performed. A secondary goal was to identify any consti-
pation care tasks performed regularly by nursing staff that were
not captured by the observational component.
Medical record review for participating residents covered
the 60 days before the date of review, which occurred within
1 week of subject observation. In addition to the resident
ORIGINAL STUDIES
descriptive data described above, the following data were
obtained: type of constipation care task performed; supplies
and medications used (amount, type, dose, etc.); preparation
and administration of special dietary supplements for constipa-
tion; date and time of performance of care task; professional
level (eg, staff nurse, CNA) of staff members performing the
task, average frequency of occurrence of each separate care task
type (by patient) and staffing required for each care task type (if
available). Task frequency data from the 60-day medical record
review was used to calculate average time per task estimate (in
units of minutes per week), average labor cost per task (dollars per
minute), average task labor cost, and average cost per task.
Data Analysis
Quantification of Costs
Time estimates for completion of each selected task were
obtained from the observational component, as were a record
of total number of staff at each professional level (staff nurse,
CNA) to obtain “staff mix” estimates per task. Staff mix
estimates were based on observational data only because the
medical record did not provide sufficient information on staff
mix by task. An average staff mix per task was estimated based
on the observational data. The average number of minutes
required for completion of each task by staff level was multi-
plied by the appropriate wage rate for each staff member, to
obtain an average labor cost per task.
The medical record abstraction component provided task
frequency estimates. After calculation of an average time per
task estimate (in units of minutes/week), the per task time
estimate was multiplied by the average labor cost per task
(dollars) to obtain an average labor cost per task. Because there
was no pay differential between day and evening shifts at the
sites, and because the majority of observation bins occurred
during morning and afternoon shifts, no shift differential costs
were calculated.
An average per task cost was calculated by adding the aver-
age per task labor cost to data obtained on the costs of
medications and supplies used in constipation care tasks. From
these values, the cost of constipation was estimated at the
level of the resident and at the level of the institution.
National wage rates were used to calculate the average cost
of care task time, using data from the Nursing Home Salary
and Benefits Report21 (effective date of data: February 1997),
to present results in terms of nationally representative US
values. Specific data on relative time involvement of staff
(nurse and CNA) was assumed to be equal for all staff. Wage
rates were inflated to 1999 values using an inflation index
provided by the Consumer Price Index for Urban Wage
Earners and Clerical Workers from the Bureau of Labor Sta-
tistics.22 Average wholesale prices of prescription and over-
the-counter medications was obtained from the Red Book.23
Prices of medical supplies were obtained from supply catalogs
and supplier websites.
Calculation of Costs
The average frequency of occurrence of each task, obtained
from averaging across all subjects, was multiplied by the av-
Frank et al. 217
Table 1. Site Information
Site 1 Site 2
Number of beds 175 334
Occupancy rate (prior 12 months) 98.3% 98.0%
Medicaid occupancy rate (prior 12 months) 34.8% 70.0%
Number of nursing staff
Charge nurse (RN) 4 17
Staff nurse 14 103
RN 9 96
LPN 5 7
Nursing aide/assistant 44 153
Number of FTEs per resident*
Charge nurse (RN) .023 .053
Staff nurse .081 .321
RN .052 .300
LPN .029 .021
Nursing aide/assistant .255 .478
Unit secretary NA† .028
Ownership For profit Non-profit
Hospital affiliation No Yes
Special care units Alzheimer’s unit Alzheimer’s unit; Sub-Acute unit; Protective
care unit; Chronic disease unit
Certification status Medicare/Medicaid Medicare/Medicaid
Number of subjects 19 40
*Number of FTEs per resident ⫽ number of full-time equivalents per position/number of current residents.
†
NA: not applicable.

erage cost for that task to obtain a “frequency adjusted” task
cost. The adjusted task costs were summed to obtain a total
estimated cost per constipated resident per year.
Average labor cost per task (dollars) ⫽ average labor time per
task ⫻ wage rate per staff level.
Average per task cost ⫽ average labor cost per task ⫹ medica-
tion costs ⫹ supply costs.
Data from the observations were entered into a computer-
ized Oracle clinical database and were double-keyed to ensure
accuracy. Data were converted to SAS and Microsoft Excel
spreadsheet formats. The unit of analysis was the individual
constipation care task. Discrepancies were resolved by con-
sultation with source documents. There were no missing data
from the observational component. Data are assumed to be
complete because it is not possible to determine what, if any,
care was not recorded in the medical record.
Sensitivity analyses were conducted on cost calculations,
first using the highest and lowest observed task time values by
task for calculation of annual cost per resident with constipa-
tion. Task frequency estimates were varied separately, to test
the effect on base case results with a frequency increase and
decrease of 50% of the mean frequency obtained from medical
chart review.
RESULTS
Site and Subject Characteristics
Descriptive information by site is presented in Table 1.
Both sites had an occupancy rate of 98% in the 12 months
before data collection. One third of residents at Site 1 and
over two thirds of the residents at Site 2 were Medicaid
funded. Site 1 had nearly half the bed capacity but less than
half the staffing of Site 2. Median staff nurse wages were
$17.07 for licensed practical nurses (LPNs) and $23.03 for
registered nurses (RNs) based on the Nursing Home Salary
and Benefits Report. Both sites had staff nurse wages above
national median values. Pay for CNAs was comparable to
national values at Site 1 ($10.40 vs. $9.67 national) but was
nearly 70% above at Site 2 ($16.41). This difference is due in
part to differences in regional wage rates. Subjects were older
on average at Site 2 (89.1 ⫾ 6.7 vs. 78.9 ⫾ 14.7, t ⫽ 3.70,
P⬍0.05) but there were no site differences on gender, race,
BMI, average number of bowel movements per week, or
MMSE score.
Subject demographic information is presented in Table 2.
Mean subject age for the sample of 59 residents was 86 (⫾11)
years; all subjects were over 65 with the exception of one. The
majority of the sample was white (97%) and female (69%).
These characteristics are comparable to those of national
nursing home data:24 88.7% white, 71.6% female. Available
MMSE scores spanned the range from lowest (0) to highest
(30). The median MDS score was closer to intact than im-
paired. Forty-six percent of subjects showed a high level of
dependence on ADLs, and the majority of subjects were on a
formal constipation preventive program. This may explain
the mean bowel movement frequency per week of nearly 7
(⫾3). The observers reported that subjects generally qualified
for enrollment based on the medication criteria, and only one
resident had ⬍2 bowel movements per week at enrollment.
Task Observation
Of the original nine constipation care tasks listed on the
data collection form, only five were observed. Constipation

218 Frank et al. JAMDA – July/August 2002

Table 2. Sample Characteristics
Characteristic N Median Minimum, Mean (SD) or
Maximum Frequency (%)
Age 59 88 45, 100 85.83 (10.97)
Gender, % female 41 — — 69.5%
Race
White 57 — — 96.6%
Black 2 — — 3.4%
Body Mass Index* 52 22.12 14.66, 35.56 22.82 (4.18)
Hospitalized during 60-day chart review period 7 — — 11.9%
(number of residents)
Average length of stay (number of days) 8 3.5 0†, 12 5.25 (4.33)
Cognitive impairment
MDS Score‡ 55 0.44 0.22, 0.89 0.51 (0.165)
MMSE Score§ 50 24 0, 30 20.04 (9.73)
ADL independence㛳
Low (score ⬍3) 30 — — 53.6%
High (score ⬎or ⫽ 3) 26 — — 46.4%
Average bowel movement frequency/week 56 6.13 1.75, 15.5 6.87 (2.84)
Bowel preventive program, % on 55 — — 93.2%
*Body Mass Index was calculated using the following formula: BMI ⫽ (weight in pounds/2.2)/(height in inches/39.37)2.
†
One patient was admitted and discharged on the same day. One patient was hospitalized twice.
‡
Minimum Data Set (MDS) cognitive status composite score where 0 ⫽ intact, 1 ⫽ impaired based on 9 MDS items: short and long term
memory; recall of season, room location, staff names/faces; presence in nursing home; easily distracted; altered perception; mental function
variability.
§
Mini-Mental State Exam (MMSE) score range 0 to 30 (higher ⫽ less impairment).
㛳
Averaged Activities of Daily Living (ADL) score from MDS data for bathing, feeding, dressing, transferring, and toileting using method of
Castle et al., 1997. Ratings range from 0: independent, 1: supervision, 2: limited assistance, 3: extensive assistance, to 4: total dependence.
Scores ⬍ 3 ⫽ “low” dependence; ⱖ 3 ⫽ “high” dependence.

evaluation was never observed, and we therefore combined
medical record review mentions of this task with the consti-
pation assessment task. Some tasks were observed in combi-
nation, precluding separate observational time estimates for
each; some of these combination tasks were assigned to a
single category. Oral medication administration observed
times may include administration of non-constipation medi-
cations as well. No additional constipation care tasks (beyond
those captured by observation) were identified through the
medical record review. A detailed explanation of specific
assumptions regarding task categorization may be obtained
from the authors.
Cost Calculations
Table 3 presents the variables that were used for calculation
of average cost per task occurrence. Median values were above
$1.25 for assessment/evaluation, administration of dietary sup-
plement, and oral medication administration. Administration
of oral medication was the most expensive task per occur-
rence, at $1.74 (low: $1.60, high: $1.89). Low and high
Nursing Home Salary and Benefits Report wage rates were
used in separate calculations to provide a national range for
costs.
Total annual labor costs per task were calculated by mul-
tiplying average task time by the frequency of task occurrence,
and multiplying that product by wage rates (See Table 4).
Supply costs were added to average labor costs to obtain the
total costs per subject per year. Total annual supply cost for
residents with constipation is nearly $400 per year, with the
majority of those costs accounted for by oral medications. All
medications provided for constipation care during the 60-day
medical record review for each subject are presented in Table
5. Based on non-PRN orders, Peri-Colace™, Colace™, sor-
bitol, and Senokot™ were the most frequently administered
medications.
The average labor cost based on staff time for oral medica-
tion administration was $1577 per resident with constipation
per year. Average labor cost for administration of dietary
supplements was more than $215 per resident with constipa-
tion per year. The labor and supply costs for all constipation
care tasks were summed to obtain the total nursing labor and
supply cost per resident with constipation per year: $2253.
Sensitivity Analyses
Specific contributors to the basic cost model were varied as
part of a sensitivity analysis. See Table 6. The lowest and
highest task times observed were input separately. For admin-
istration of dietary supplement, the lowest task time decreased
the base case cost estimate by approximately $6, and the
highest time increased it by nearly $600. The lower impact of
use of the low time estimate is most likely because most
observations (70%) were below the mean value used in the
base estimate. Varying time estimates for oral medication
administration had the largest effect on the total cost esti-
mate. Using the lowest observed time for oral medication
administration in the calculation (5 seconds) reduced the cost
estimate by more than $1100. Use of the highest time esti-
mate (more than 11 minutes) increased the estimate by over
$2400. Results were similar when task frequency rather than
task time was altered, using ⫹50% and ⫺50% of mean time

ORIGINAL STUDIES Frank et al. 219

 based on medical chart review data. Again, the biggest effect

1997–1998 Nursing Home Salary and Benefits Report converted to present wages by Consumer Price Index for Medical Care Services (10/20/99) as wage source. Wages calculated by increasing
$1.16

$1.24

$0.66
$0.90
$1.60
Low
was found for oral medication administration: ⫺$1066 and
⫹$1058 for the ⫺50% and ⫹50% frequency estimates, re-
Task Labor Cost per spectively. For all other constipation care tasks, the differ-

Median
ences were within $100.

$1.28

$1.38

$0.72
$0.97
$1.74
DISCUSSION
Occurrence
(Dollars)†

an hourly rate by 30% to include benefits (Hospital and Healthcare Compensation Service & John R. Zabka Associates, Inc, 1997). Effective date of data: February 1997.
This is the first study of its kind to use an empirical
$1.41

$1.51

$0.81
$1.05
$1.89
approach to examine the nursing labor and supply costs asso-
High

ciated with constipation care in nursing homes. Despite the

high prevalence of constipation care in nursing homes, avail-
able empirical data on it are extremely limited. This study
Average Task
Time (mm:ss)

demonstrated that the per patient cost of constipation in

nursing homes is considerable.
3:24

3:38

4:29
2:35
4:37
These results are relevant to cost management and care
quality in long-term care. There is a growing literature on
nurse occupational stress. Although certain aspects of nursing
care in long-term care are immutable (eg, regulations requir-
Occurrences

ing certain staff to administer medications, for example),

Number of
Observed

making positive changes to improve staff time use can be

20

10

2
1
182

associated with improved morale, and effective use of nursing

Total

staff time is related to better care quality in long-term

care.25,26 Nursing assistants comprise 45% to 70% of long
term care employees in North America27 and provide up to
$20.56

$20.53

$20.81
$20.72

90% of personal care in nursing homes.28 Improving care

Average Labor Cost (Wage in

$8.82
Low

efficiency for common care tasks like constipation care may

Dollars/Hour Per Nurse)†

assist with reducing staff stress and burnout.

The dominant contribution of oral medication administra-
Median

tion and dietary supplement administration suggests that

$22.69

$22.70

$22.62
$22.65
$9.67

these two tasks are promising areas in which to design cost-

saving interventions. The dietary supplementation observed
for this study related to constipation only; supplementation
for nutritional needs was not included. Therefore, treatments
$24.88

$24.94

$10.81
$24.30
$24.51
High

or intervention strategies that reduce or eliminate the need

for constipation-related supplementation could result in sub-
stantial savings in terms of staff time and labor costs. Because
CNA*

of the high rate of polypharmacy among residents, the oral

medication administrations observed here often included
0

2
0
0

other non-constipation drugs. Therefore, not all instances of

oral medication administration can be eliminated by reducing
Average Staff Mix

Nurse
Staff

administration of constipation-related medications. However,

Calculation of Task Labor Costs per Occurrence

155
11

reducing time to administer the constipation medication (eg,

5

0
1

by reducing number of doses, time needed for pill crushing, or

time needed for assisting residents) could have a notable cost
Charge
Nurse

impact. Constipation treatments with better efficacy could

30

also lower care costs by eliminating the need for multiple

9

0
0

constipation medications (eg, maintenance dosing of milk of

magnesia multiple times per day) and/or eliminating the need
for dietary supplementation distinct from regular mealtimes.
Suppository administration

*CNA: Certified Nurse Aide.

Assessment/evaluation of

Because some medication administrations will occur whether

constipation medications are included are not, the cost im-
Enema administration
(prune/apple/bran
Dietary supplement

pact of reducing or eliminating the need for constipation-

administration

administration

specific medications is limited. A broad range for sensitivity

Oral medication
supplement)
constipation

analysis was chosen to ensure conservative cost estimation.

Swallowing disorders and other eating-related difficulties are
Table 3.

common among residents, and nurses’ time to accommodate

these problems during multiple daily medication administra-
Task

tions may account for much of the observed time.

220 Frank et al. JAMDA – July/August 2002

Table 4. Labor and Supply Costs by Task and Total Annual Task Cost
Task Average Supply Cost* (Per Labor Cost Frequency (Per
Task Time Subject per Year) (Subject per Year) Subject per Year)
(mm:ss) NHSBR† Median
Assessment/evaluation of constipation 3:24 $0.00 $38.97 30.33
Dietary supplement administration 3:38 $10.59 $214.70 148.37
(prune/apple/bran supplement)
Enema administration 4:29 $6.56 $5.53 7.64
Suppository administration 2:35 $9.57 $16.98 17.44
Oral medication administration 4:37 $372.63 $1577.21 904.87
Total annual cost per resident with — $399.35 $1853.40 —
constipation by component
Total annual cost (labor ⫹ supplies) per resident with constipation: $2252.75†
*Medication average wholesale price obtained from Red Book (1999). Medical supply prices were obtained from supply catalogs and supplier
websites.
†
1997–1998 Nursing Home Salary and Benefits Report converted to present wages by Consumer Price Index for Medical Care Services
(10/20/99) as wage source. Wages calculated by increasing an hourly rate by 30% to include benefits (Hospital and Healthcare Compensation
Service & John R. Zabka Associates, Inc. 1997). Effective date of data: February 1997.
Total annual cost ⫽ average time per task ⫻ (60-day frequency chart/59 residents)/60 * 365.25.

It was of interest that frequency of bowel movement was
not severely limited in this sample, and that subject eligibility
was below literature-based prevalence estimates. The fluctu-
ating symptom expression of constipation and our limited
enrollment window may, in part, account for this finding. The
bowel programs in place for more than 90% of the subjects,
which specify close monitoring of bowel movement frequency
and treatment if ⬍1 movement occurs per 3 days, most likely
account for the bowel movement frequency findings. These
protocols represent standard care, and therefore, we do not
think that they resulted in more or less care than is generally
given to patients identified with constipation.

Table 5. Medications, Total Number of Administrations, and Average Cost per Dose for Constipation Care
Drug Name Average Number of Administrations Average Cost
per Patient per Year* per Dose ($)*
Peri-Colace姞 256.70 0.42
Colace姞 191.60 0.41
Sorbitol 171.17 0.15
Senokot姞 116.18 0.35
Dulcolax姞 (oral) 39.41 0.68
Phillips Milk of Magnesia 38.69 0.10
Lactulose 37.56 1.71
Metamucil姞 22.91 0.15
Surfak姞 18.57 0.31
Dulcolax姞 (suppository) 10.42 0.68
Cascara 6.19 0.49
Bisacodyl (oral) 5.88 0.01
Bisacodyl (suppository) 0.21 0.09
*Average number of administrations per year ⫽ total number of administrations for 60-day retrospective medical record review (3540
patient days) ⫻ 365.25/3540.

Table 6. Sensitivity Analyses*

Task Observed Time Frequency of Occurrence
Lowest Highest ⴚ50% ⴙ50%
Assessment/evaluation of constipation NA NA $2229.53 $2268.50
Dietary supplementation $2246.59 $2851.17 $2228.60 $2269.43
Oral medication administration $1151.39 $4658.84 $1187.01 $3311.03
Enema administration $2246.70 $2251.35 $2239.70 $2258.33
Suppository administration NA† NA $2235.74 $2262.29
*Base case value: $2252.75.
†
Task time was not included for suppository administration since there was a single observation point.

ORIGINAL STUDIES Frank et al. 221

 It is possible that patients are medicated after the symptoms
of constipation subside. However, the chronic nature of con-
stipation in this patient population suggests that medication
may be classified as maintenance for many patients after an
acute treatment period, with treatment titrated to address
fluctuating symptoms. Symptoms were not always document-
ed; therefore, we did not attempt to document the numbers
with symptoms but instead made the assumption that treat-
ment indicated some evidence of symptoms sufficient to war-
rant treatment in the staff’s judgment.
The low number of observed occurrences of enema and
suppository administration (fewer than 10) limits the inter-
pretation of results based on those data. However, observed
completion times for these tasks were not out of scale based
on data from the other tasks. Focused observation of these
tasks and/or extension of the observation schedule could
improve confidence in these estimates.
Of our original list of nine tasks, the four tasks that we did
not observe were bowel movement toileting, fecal disimpac-
tion, evaluating for response to treatment for constipation,
and tasks that result from fecal incontinence. Evaluation for
treatment response was mentioned in the medical record and
was, for observational purposes, indistinguishable from assess-
ment/evaluation of constipation. We therefore included all
evaluations as part of the assessment/evaluation task rather
than try to distinguish them. In this way, staff time and
frequency of occurrence were still meaningfully captured. Fe-
cal disimpaction was always combined with enema adminis-
tration, and therefore, the separate category was dropped. No
instances of fecal incontinence were observed for the residents
during our observation times. Staff indicated that, because of
constipation treatment, fecal incontinence is a relatively in-
frequent occurrence, and we were therefore not surprised that
we did not observe it. Finally, we did not observe bowel
movement toileting. Based on chart review, it seems that such
toileting occurred very infrequently, and our observation bins
were not long enough to capture this relatively infrequent task
completion.
Data collectors noted that constipation-related care was
being provided to residents who were not enrolled in the
study (because they did not meet inclusion criteria). Our cost
estimates are therefore likely conservative. The per patient
annual costs provided by this study can be used with a range
of prevalence estimates to obtain a broader range within
which most facility costs are likely to fall.
The per patient annual cost of constipation estimated from
this study was approximately $2253. It is helpful to under-
stand the cost of constipation care in nursing homes in rela-
tion to the costs of other common disorders in that setting.
The literature on UI was used as a basis for comparison
because multiple cost studies were available. Based on work by
Wagner and Hu,29 the annual cost for behavioral and phar-
macologic treatment and routine care per UI patient (inflated
to 1999 dollars), is $664. In a separate report, Hu et al.30
estimated labor costs for residents with UI without a catheter
at $6/day per resident. Inflating to 1999 dollars, that results in
a per patient annual cost of $3317.30. The per patient annual
cost of constipation estimated from the results reported here
222 Frank et al.
was $2253, far higher than the low estimate for UI treatment
of $664 and closer to the high UI estimate of $3317. Given
the extensive nursing home UI cost literature and scarce
constipation literature, and in light of the high cost of con-
stipation found in this study, additional research resources
should be devoted to studying the costs associated with con-
stipation in the nursing home.
Limitations
These data provide an average cost per episode, but the true
cost for constipation care cannot be known until better prev-
alence estimates are obtained. Because multiple medications
were frequently administered at a single time, obtaining a
separate cost for oral constipation medication administration
was not feasible, and therefore the extent to which reducing
oral constipation medication can reduce costs requires addi-
tional study. Because identification of constipation-only med-
ication administration was not possible, the marginal cost of a
specific class of medication may be minimal. Clearly further
research is warranted.
The frequency estimates for constipation care tasks were
based on medical record review. The inaccuracies in medical
records are widely acknowledged, and quantifying the true
task frequency requires direct observation beyond that per-
mitted by this study design. Given our observation schedule,
results can be generalized to daytime care, that is, that pro-
vided between 7:00 AM and 7:00 PM. Finally, the lack of
observational data on toileting suggest that these cost esti-
mates are likely conservative.
The extent to which findings from these homes can be
generalized nationally may be limited. Despite similarities in
resident demographic characteristics, staff nurse wages were
above national medians at both sites, and at Site 2, CNAs
were paid 70% above the national median wage rate. Also,
Site 2 had a higher level of staffing, in terms of staff-to-
resident ratio, than Site 1. Informal observation by study staff
suggested that this site was notably superior, in terms of
organization and appearance, to other nursing homes.
Finally, a conceptual distinction can be made between
constipation prevention and constipation treatment. How-
ever, in practice, prevention and treatment are difficult to
distinguish. Constipation prevention includes PRN medica-
tions and non-PRN medications. The treatment versus pre-
vention distinction is, therefore, difficult to make with a
disorder like constipation, in which treatment is generally
chronic with adjustments for acute symptoms. In this study,
most subjects identified by nursing staff as constipated were in
fact on a treatment regimen. Therefore, the data most likely
reflect the cost of constipation treatment in nursing homes
rather than the costs associated with just prevention.
CONCLUSION
Identifying areas in which nursing care can be improved, in
terms of time efficiency and effectiveness, is critical to ad-
dressing nursing home care quality. The prospective payment
system (PPS) creates an incentive for nursing home admin-
istrators to improve care efficiency in terms of staff time. For
Medicare PPS payments, for example, health care institutions
JAMDA – July/August 2002
that operate efficiently benefit if actual costs are below the
reimbursed payment level.31 In addition to the financial ben-
efit, reducing nursing staff time required for one treatment
category can allow nursing staff to spend more time providing
quality patient care. Constipation is a therapeutic area with
substantial consequences for nursing homes. Identifying con-
tributors to care costs in general and nursing staff time in
particular can assist policymakers, administrators, and staff
with improving care in nursing homes.
ACKNOWLEDGMENTS
The authors thank Emuella Flood, Sonika Mathur, and
Sarah Nason of MEDTAP, and Linda Armstrong, Suzanne
Beckner, Helen Leahy, Shirley Lewis, and Katherine Muth of
WESTAT for their assistance. In addition, we are grateful to
the residents and staff of the participating nursing homes for
their participation.
REFERENCES
1. Stewart RB, Moore MT, Marks RG, Hale WE. Correlates of constipation
in an ambulatory elderly population. Am J Gastroenterol 1992;87:859 –
864.
2. Talley NJ, Fleming KC, Evans JM, O’Keefe EA, et al. Constipation in an
elderly community: A study of prevalence and potential risk factors.
Am J Gastroenterol 1996;91:19 –25.
3. Alessi CA, Henderson CT. Constipation and fecal impaction in the
long-term care patient. Clin Geriatr Med 1988;4:571–588.
4. Williams SG, DiPalma JA. Constipation in the long-term care facility.
Gastroenterol Nurs 1990;12:179 –182.
5. Romero Y, Evans JM, Fleming KC, Phillips SF. Constipation and fecal
incontinence in the elderly population. Mayo Clin Proc 1996;71:81–92.
6. Donald IP, Smith RG, Cruikshank JG, et al. A study of constipation in
the elderly living at home. Gerontology. 1985;31:112–118.
7. Fonda D, Woodward M, D’Astoli M, Chin WF. Sustained improvement
of subjective quality of life in older community-dwelling people after
treatment of urinary incontinence. Age Ageing 1995;24:283–286.
8. Borrie MJ, Davidson HA. Incontinence in institutions: Costs and con-
tributing factors. Can Med Assoc J. 1992;147:322–328.
9. Frank L, Kleinman L, Schmier J, Farup C. The burden of constipation
care in nursing homes. Presented to the Gerontological Society of Amer-
ica, November, 1998. Philadelphia, PA. Published abstract. Gerontolo-
gist 1998;38 (Special Issue):194 –195.
10. Sonnenberg A, Everhart JE, Brown DM. The economic cost of consti-
pation. In: Kamm MA, Lennard-Jones JE, eds. Constipation. Bristol, PA:
Wrightson Biomedical Pub Ltd., 1994, pp. 19 –29.
11. Potter DEB. Design and methods of the Medical Expenditure Panel
Survey Nursing Home Component. Rockville (MD): Agency for Health
Care Policy and Research. MEPS Methodology Report 1998; No. 3.
AHCPR Pub. No. 98 – 0041.
ORIGINAL STUDIES
12. Thompson WG, Creed F, Drossman DA, et al. Functional bowel disor-
ders and chronic functional abdominal pain. Gastroenterol Int 1992;5:
75–91.
13. Hawes C, Morris JN, Phillips CD, et al. Reliability estimates for the
Minimum Data Set for nursing home resident assessment and care
screening (MDS). Gerontologist 1995;35:172–178.
14. Morris JN, Hawes C, Fries BE, et al. Designing the National Resident
Assessment Instrument for Nursing Homes. Gerontologist 1990;30:293–
307.
15. Hartmaier SL, Sloane PD, Guess HA, Koch GG. The MDS Cognition
Scale: A valid instrument for identifying and staging nursing home
residents with dementia using the Minimum Data Set. J Am Geriatr Soc
1994;42:1173–1179.
16. Castle NG, Fogel B, Mor V. Risk factors for physical restraint use in
nursing homes: Pre- and post-implementation of the Nursing Home
Reform Act. Gerontologist 1997;37:737–747.
17. Cella M. The nursing costs of urinary incontinence in a nursing home
population. Nurs Clin North Am 1988;23:159 –168.
18. Cummings V, Holt R, van der Sloot C, et al. Costs and management of
urinary incontinence in long-term care. J Wound Ostomy Continence
Nurs 1995;22:193–198.
19. Schnelle JF, Keeler E, Hays RD, et al. A cost and value analysis of two
interventions with incontinent nursing home residents. J Am Geriatr
Soc. 1995;43:1112–1117.
20. Sowell VA, Schnelle JF, Hu TW, Traughber B. A cost comparison of five
methods of managing urinary incontinence. QRB Qual Rev Bull. 1987;
13:411– 414.
21. Hospital and Healthcare Compensation Service, Oakland, NJ: John R.
Zabka Associates, Inc. Published May; Copyright 1997 by John R. Zabka
Associates, Inc 1997.
22. Bureau of Labor Statistics. Most Requested Series page: Consumer Price
Index for Urban Wage Earners and Clerical Workers. Available at:
http://www.bls.gov/data/top20.html. Accessed May 23, 2002.
23. 1999 Red Book. Montvale NJ: Medical Economics Company, 1999.
24. Krauss NA, Altman BM. Characteristics of nursing home residents–
1996. Rockville, MD: Agency for Health Care Policy and Research.
MEPS Research Findings, No.5. AHCPR Pub. No. 99 – 0006, 1998.
25. Cardona P, Tappen RM, Terrill M, et al. Nursing staff time allocation in
long-term care. A work sampling study. J Nurs Adm 1997;27:28 –36.
26. Burgio LD, Engel BT. Hawkins A, et al. A descriptive analysis of nursing
staff behaviors in a teaching nursing home: differences among NAs,
LPNs and RNs. Gerontologist. 1990;30:107–112.
27. Novack M, Chappell NL. Nursing assistant burnout and the cognitively
impaired elderly. Int J Aging Hum Dev 1994;39:105–120.
28. Chappell NL, Novak M. The role of support in alleviating stress among
nursing assistants. Gerontologist. 1992;32:351–359.
29. Wagner TH, Hu TW. Economic costs of urinary incontinence in 1995.
Urology 1998;51:355–361.
30. Hu TW, Kaltreider DL, Igou JF, et al. Cost effectiveness of training
incontinent elderly in nursing homes: A randomized clinical trial. Health
Serv Res 1990;25:455– 477.
31. McMahon T, O’Brien T, de Lissovoy G. Health care prospective
payment: The United States experience. Bethesda: MEDTAP,
2000.
Frank et al. 223