Productivity

1
Chapter 9. Productivity
2
Outline

• Trends in Healthcare Productivity Consequences
  of PPS
• Productivity Definitions and Measurements
• Productivity Benchmarking
• Multifactor Productivity
• Commonly Used Productivity Ratios
• Hours per Patient Day or Visit
• Adjustment for Inputs
• Skill-Mix Adjustment to Worked Hours
• Cost of Labor
• Adjustments for Output Measures
• Service/Case-Mix Adjustments
• Productivity Measures Using Direct Care Hours
• Productivity Quality Relationship
• Productivity Dilemmas
• Multiple Dimensions of Productivity New Methods
• Data Envelopment Analysis (DEA)
• Productivity Improvement

3
Trends in Productivity Consequences of PPS

• The recent decades changes in reimbursement
  strategies aimed to end waste and promote
  innovative and cost-efficient delivery systems.
• productivity gains from PPS have not materialized
  to the extent predicted.
• Hospitals now employ more people to treat fewer
  patients, and the increase is not accounted for
  by the greater severity of patient illness in the
  late 1980s and in1990s.
• Although employers, insurers and public are
  spending less on inpatient care, the rising use
  of outpatient procedures has simply increased
  costs in that area which counters the savings
  (Altman, Goldberger, and Crane, 1990).

4
Trends in Productivity Consequences of PPS

• The constraints that force healthcare
  institutions into the role of cost centers,
  coupled with shifting patterns of inpatient
  acuity, tight healthcare labor markets, and
  society's expectations of high quality of care
  are leading healthcare organizations to a
  "productivity wall." When the wall is reached,
  it is quality of care that inevitably is
  sacrificed for the sake of productivity and
  profit (Kirk, 1990).
• It must be recognized that there are limits to
  ratcheting up productivity.
• It is not always possible to do more with less.

5
Productivity Definitions and Measurements

• Productivity is one measure of the effective use
  of resources within an organization, industry, or
  nation.
• The classical productivity definition measures
  outputs relative to the inputs needed to produce
  them. That is, productivity is defined as the
  number of output units per unit of input

6
Productivity Definitions and Measurements

• Sometimes, an inverse calculation is used that
  measures inputs per unit of output. Care must be
  taken to interpret this inverse calculation
  appropriately the greater the number of units of
  input per unit of output, the lower the
  productivity.
• For example, traditionally productivity in
  hospital nursing units has been measured by hours
  per patient day (HPPD). That requires an
  inversion of the typical calculations meaning
  total hours are divided by total patient days.

7
Example 9.1
Nurses in Unit A worked collectively a total of
25 hours to treat a patient who stayed 5 days,
and nurses in Unit B worked a total of 16 hours
to treat a patient who stayed 4 days. Calculate
which of the two similar hospital nursing units
is more productive.
Solution
First, define the inputs and the outputs for the
analysis. Is the proper measure of inputs the
number of nurses or of hours worked? In this
case the definition of the input would be total
nursing hours. When the total number of nursing
hours worked per nurse is used as the input
measure, then the productivity measures for the
two units are

8
Productivity Definitions and Measurements

• Productivity Benchmarking. Productivity must be
  considered as a relative measure the calculated
  ratio should be either compared to a similar
  unit, or compared to the productivity ratio of
  the same unit in previous years. Such
  comparisons characterize benchmarking. Many
  organizations use benchmarking to help set the
  direction for change.
• Historical Benchmarking is monitoring an
  operational units own productivity or
  performance over the last few years. Another way
  of benchmarking is to identify the best practices
  (best productivity ratios of similar units)
  across health organizations and incorporate them
  in ones own.

9
Productivity Definitions and Measurements
Multifactor Productivity. Example 9.1
demonstrated a measure of labor productivity.
Because it looks at only one input, nursing
hours, it is example of a partial productivity
measure. Looking only at labor productivity may
not yield an accurate picture.
Newer productivity measures tend to include not
only labor inputs, but the other operating costs
for the product or service as well.
10
Example 9.2
A specialty laboratory performs lab tests for the
area hospitals. During its first two years of
operation the following measurements were
gathered Measurement Year 1 Year 2 Price
per test () 50 50 Annual
tests 10,000 10,700 Total labor
costs() 150,000 158,000 Material costs ()
8,000 8,400 Overhead () 12,000
12,200 Determine and compare the multifactor
productivity for historical benchmarking.
.


Solution
11
Commonly Used Productivity Ratios

• Hours Per Patient Day (or Visit)

inpatient
outpatient
12
Commonly Used Productivity Ratios
Example 9.3
Annual statistical data for two nursing units in
Memorial Hospital are as follows Measurements U
nit A Unit B Annual Patient Days 14,000
10,000 Annual Hours Worked 210,000 180,000 C
alculate and compare hours per patient day for
two units of this hospital.
Solution
hours
hours
13
Commonly Used Productivity Ratios
Example 9.4
Performsbetter Associates a two-site group
practice, requires productivity monitoring. The
following initial data are provided for both
sites of the practice Measurements
Suburban Downtown Annual Visits 135,000
97,000 Annual Paid Hours 115,000 112,000 Calcu
late and compare the hours per patient visit for
the suburban and the downtown locations of this
practice.
Solution
hours or 51 minutes.
hours or 69 minutes.
14
Adjustments for Inputs
Skill-Mix Adjustment weigh the hours of personnel
of different skill levels by their economic
valuation. One approach is to calculate weights
based on the average wage or salary of each skill
class. To do that, a given skill class
wage/salary would be divided into the top class
skill salary. If RNs, LPNs and Aides are
earning 35.00, 28.00, and 17.50 an hour,
respectively Then, one hour of a nurse aides
time is economically equivalent to 0.5 hours of a
RN's time and one hour of a LPN's time is equal
to 0.8 hours of a RN's time.
.
15
Adjustments for Inputs
Adjusted Hours 1.0(RN hours) 0.8(LPN hours)
0.5(Aide hours)
16
Adjustments for Inputs
Adjusted Hours 1.0(RN hours) 0.8(LPN hours)
0.5(Aide hours)
17
Adjustments for Inputs
Similarly, in outpatient settings, if one hour of
a nurse practitioner's (NP) time is economically
equivalent to 0.6 hours of a specialist's (SP)
time, and if one hour of a general practitioners
(GP) time is equal to 0.85 hours of a
specialists time, adjusted hours would be
calculated as
.
Adjusted Hours 1.0 (SP hours) 0.85 (GP hours)
0.6 (NP hours)
18
Adjustments for Inputs
Example 9.5 Using data from Example 9.3, and
economic equivalencies of 0.5 Aide RN, 0.8 LPN
RN, calculate the adjusted hours per patient
day for Unit A and Unit B. Unit A at Memorial
Hospital employs 100 RNs. The current skill
mix distribution of Unit B is 45 RNs, 30 LPNs,
and 25 nursing aides (NAs). Compare
unadjusted and adjusted productivity scores.
19
Adjustments for Inputs
Solution The first step is to calculate
adjusted hours for each unit. For Unit A, since
it employs 100 RNs, there is no need for
adjustment. For Unit B Adjusted Hours (Unit
B) 1.0 (180,000.45) 0.80 (180,000.30)
0.50 (180,000.25). Adjusted Hours (Unit B) 1.0
(81,000) 0.80 (54,000) 0.50
(45,000). Adjusted Hours (Unit B) 146,700. In
this way, using the economic equivalencies of the
skill-mix, the number of hours is standardized as
146,700 instead of 180,000.
Standardized Cost of Labor.
hours.
hours.
Using adjusted hours, Unit A, which appeared
productive according to the first measure (see
example 9.3), no longer appears as productive.
20
Adjustments for Inputs
Standardized Cost of Labor. Total labor cost
comprises the payments to various professionals
at varying skills. To account for differences in
salary structure across hospitals or group
practices, cost calculations can be standardized
using a standard salary per hour for each of the
skill levels
.
Labor Cost RN wages (RN hours)
LPN wages (LPN hours)
NA wages (Aide hours).
21
Adjustments for Inputs
Example 9.6 Performsbetter Associates in
Example 9.4 pays 110, 85, and 45 per hour,
respectively, to its SPs, GPs and NPs in both
locations. Currently, the suburban location
staff comprises of 50 SPs, 30 GPs, and 20 NPs.
The downtown location, on the other hand,
comprises 30 SPs, 50 GPs, and 20 NPs.
Calculate and compare the labor cost of care,
and labor cost per visit for both locations.
22
Adjustments for Inputs
Solution
First, calculate Labor Cost of Care for each
location. Labor Cost SP wages (SP hours) GP
wages (GP hours) NP wages (NP hours), Labor
CostSuburban 110 (115,0000.50) 85
(115,0000.30) 45 (115,0000.20). Labor
CostSuburban 110 (57,500) 85 (34,500) 45
(23,000). Labor CostSuburban 10,292,500. Labor
CostDowntown 110 (112,000.30) 85
(112,0000.50) 45 (112,0000.20). Labor
CostDowntown 110 (33,600) 85 (56,000) 45
(22,400). Labor CostDowntown 9,464,000.
23
Adjustments for Outputs
Service-Mix Adjustments. Service-mix adjustment
is useful tool for comparison of, for instance,
two community hospitals that provide different
services or have significantly different
distributions of patients among their services.
The service-mix adjusted volume is weighted by a
normalized service-intensity factor.
.
24
Adjustments for Outputs
Service-Mix Adjustments
.
Example 9.7 Two hospitals, each with
unadjusted volume of 10,000 patient days per
month, provide only two services, S1 and S2,
requiring respectively 3 and 7 hours of nursing
time per patient day. Hospital A has a
service-mix distribution of 2000 patient days for
S1 and 8000 patient days for S2. Hospital B has
8000 days for S1 and 2000 days for S2.
Calculate adjusted patient days for both
hospitals.
25
Adjustments for Outputs
Service-Mix Adjustments
Solution In this case, total unadjusted volume
is simply the sum of the volume for each service
in each hospital, or Unadjusted Volume X1 X2.
Hospital-A Hospital-B
Service S1 (3 hours/patient day) X12000 X18000
Service S2 (7 hours/patient day) X28000 X22000
Total Unadjusted Volume 10,000 10,000
.
Adjusted Volume W1X1 W2X2.
Adjusted volume for Hospital-A
0.62,0001.48,000 12,400. Adjusted volume for
Hospital-B 0.68,0001.42,000 7,600.
26
Adjustments for Outputs
Case-Mix Adjustments. The methodology for
case-mix adjustment is similar to that for
service-mix adjustment. Although most hospitals
rely on advanced acuity systems, each system is
based on the weight factors for the different
acuity categories. Patients in each category
require similar amounts of nursing care over a
given 24 hour time period however, across
categories the care requirements differ
significantly. For acuity, the focus is on
patients direct care requirements. The ratio
of the hours of direct care provided to the total
hours worked is another measure of productivity.
27
Adjustments for Outputs
Case-Mix Adjustments
Example 9.8 Unit A and Unit B (from Example
9.3), a medical care unit in Memorial Hospital,
classify patients into four acuity categories
(Type I through Type IV), with direct care
requirements per patient day being respectively,
0.5, 1.5, 4.5, and 6.0 hours. Annual
distributions of patients in these four acuity
categories in Unit A were 0.15, 0.25, 0.35, and
0.25. Annual distributions of patients in Unit
B were 0.15, 0.30, 0.40, and 0.15. Calculate
the case mix for these two units, and determine
which unit has been serving more severe
patients.
28
Adjustments for Outputs
Case-Mix Adjustments
Solution
.
.
.
.
.
.
29
Adjustments for Outputs
Case-Mix Adjustments
Once the case-mix is determined, the output side
of the productivity ratios can be adjusted by
simply multiplying volume (patient days,
discharges, visits) by case-mix index
as Adjusted Patient Days Patient Days
Case-mix index. Adjusted Discharges Discharges
Case-mix index. Adjusted Visits Visits
Case-mix index.
30
Productivity Measures Using Direct Care Hours
Hours of Direct Care. Hours of direct care is
an important component of productivity ratios. It
serves as a building block for other ratios.
To illustrate its development, let us assume
that patients are categorized into acuity
groupings requiring H1, H2, H3, ., Hm hours of
direct nursing care per patient day. Further,
assume that there are N1, N2, N3, ., Nm annual
patient days in units 1 through m. The total
amount of direct nursing care in nursing unit j
would be calculated as
31
Productivity Measures Using Direct Care Hours
Percentage of Hours in Direct Care. This is an
additional measure can be derived from the Hours
of Direct Care calculation, as the ratio of
direct care hours to total care hours.
Percentage of Adjusted Hours in Direct Care. We
also can determine the percentage of adjusted
nursing hours as adjusted for skill-mix in direct
patient care.
32
Productivity Measures Using Direct Care Hours

• Example 9.9
• Using information from Examples 9.3 and 9.8
• calculate
• hours of direct care
• percentage of hours in direct care, and
• percentage of adjusted hours in direct care
• for Units A and B of Memorial Hospital.
• Compare these results in terms of percentage of
• adjusted hours in direct care.

33
Productivity Measures Using Direct Care Hours
Solution Memorial Hospital uses an acuity
classification system with 4 categories of direct
hours of care per patient day 0.5, 1.5, 4.0, and
6.0 hours. The annual distributions of patients
in these four acuity categories in Unit A were
0.15, 0.25, 0.35, and 0.25. The annual
distributions of patients in Unit B were 0.15,
0.30, 0.40, and 0.15. Annual patient days for
Unit A were 14,000, and for unit B 10,000.
Annual hours worked were 115,000 and 112,000,
respectively.
34
Productivity Measures Using Direct Care Hours
Solution
.
.
35
Productivity Measures Using Direct Care Hours
Solution
.


36
Figure 9.1 Productivity and Quality Tradeoff
A
Q
QA
A
A
B
I
IA
Source Shukla, R.K. Theories and Strategies of
Healthcare Technology-Strategy-Performance,
Chapter 4, Unpublished Manuscript, 1991.
Printed with permission.
37
Productivity Wall?

• Quality is difficult to measure, and its
  definition is ambiguous
• The relationships between quantity of care
  provided and quality are often uncertain

38
Many people confuse. . .

• The concepts of productivity, efficiency, and
  effectiveness.

39
Its quite simple really!

• Efficiency-- using the minimum number of inputs
  for a given number of outputs
• Effectiveness-- refers to outputs are the
  proper inputs being used to produce the
  appropriate outcomes?
• Productivity-- a broader concept than
  efficiency refers to effective use of a given
  set of resources

40
But efficiency has varying dimensions..

• Technical Efficiency-- relationship between
  various inputs and related outputs use minimum
  combination of resources for a given level of
  quantity or level of care.
• Allocative (Economic) efficiency-- adds cost to
  the measure of technical efficiency.

41
Graphically,
Iso-cost
Assume NPs and MDs can be substituted. The
hospital can either use 3 MDs and 2 NPs (pt. A),
or 1 MD and 5 NPs (pt. B). Both result in the
same level of quality and can produce the
same quantity of output.
Isoquant
Are points A and B both technically efficient? Is
point C technically efficient, why or why
not? Remember what an isoquant is? Are all
points on an isoquant technically efficient?
economically efficient?
42
Lets expand our discussion. . .

• Data envelopment analysis is a recently developed
  technique that can be used to measure the
  multiple dimensions of productivity.
• It allows multiple inputs and outputs to be used
  in a linear programming model that develops a
  score of technical efficiency.

43
Data Envelopment Analysis (DEA)

• DEA can be used to measure productivity of
  hospitals, physicians, group practices, or any
  other unit of analysis, referred to as the
  decision making unit (DMU)
• The technical efficiency score of optimally
  producing DMUs equals 1 (and lies on the
  isoquant). All other DMUs are measured against
  these technically efficient DMUs, and have a
  score of between 0 and 1.

44
DEA-- A Simple Example
Inefficiency
Supplies
Physicians P1, P2, and P3 are technically
efficient, ceteris paribus, and would receive
an efficiency score of 1. Physician 4, however
is inefficient and must reduce either visits
and or use of medications to become as efficient
as his/her peers. The amount of the reduction
necessary is called inefficiency.
4 3 2 1
P2
P4
P1
P3
LOS
0 1 2 3
45
DEA-- An ApplicationOzcan and Luke (1993), A
National Study of the Efficiency of Hospitals in
Urban Markets

• The study examines the contribution of various
  hospital characteristics to hospital technical
  efficiency
• Outputs included
• Treated cases
• Outpatient visits
• Teaching FTEs
• Inputs included
• Capital
• Plant complexity
• Labor
• Supplies

46
DEA Applications, cont.

• Slack values allow the manager to determine just
  how much the input/output mix must be changed for
  inefficient DMUs to reach efficiency
• DEA is also useful for benchmarking or
  development of report cards, making it
  particularly useful in a managed care environment

47
Improving Healthcare Productivity

• Develop productivity measures for all operations
  in their organization,
• Look at the system as a whole (do not
  sub-optimize) in deciding on which
  operations/procedures to focus productivity
  improvements.
• 3. Develop methods for achieving productivity
  improvements, and especially benchmarking by
  studying peer healthcare providers that have
  increased productivity and reengineer care
  delivery and business processes.
• 4. Establish reasonable and attainable standards
  and improvement goals.
• 5. Consider incentives to reward workers for
  contributions and to demonstrate managements
  support of productivity improvements.
• 6. Measure and publicize improvements.

48
The End