VBHC - week 3 Flashcards

1
Q

Relevance of cost measurement in VBHC

A
  • Required for completing the value equation and thus for comparing, steering on, and improving overall value
    o Improving outcomes cannot come at any cost
  • Provides insight in the process and cost of care delivery, which in turn helps to:
    o Inform focused process improvements and care redesign
    o Appropriately target cost reduction efforts
    o Meet external expectations and pressure
  • Informs the design of (value-based) payment strategies
    o E.g. estimating the cost of the full care cycle
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

why tdabc

A
  • Used in other industries since the 1990s as a simple and flexible costing method to replace the resource-demanding ABC-method
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q
  • Main differences with traditional ABC are that TDABC (Kaplan & Anderson 2004):
A
  1. Uses patients and their conditions as unit of analysis
    o Not departments, procedures or services
  2. Acknowledges that resources will not work at full capacity
    o Practical vs. theoretical capacity
  3. Focuses on the time it takes to complete a step in a care process
    o Not the % of time employees expect to spend on activities
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

the seven steps of TDABC

A
  1. Select the medical condition
  2. Define the “care delivery value chain” (CDVC)
  3. Develop process maps for each element in the CDVC
  4. Obtain time estimates for each step/activity in the process maps
  5. Estimate the cost of resources involved
  6. Estimate the capacity of each resource & calculate its capacity cost rate
  7. Calculate the total cost of care per patient
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q
  1. Select the medical condition
A

“An interrelated set of patient circumstances that are best addressed in a coordinated way and should be broadly defined to include complications and comorbidities.” (Kaplan & Porter 2011)
* Define the start and end of the care cycle for the condition

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q
  1. Define the care delivery value chain (CDVC = care delivery value chain)
A
  • Provides an overall view of care delivery by charting all main activities and their locations occurring over the care cycle
  • Ideally disregards boundaries between departments, specialties, organizations, etc. for the patients this doesn’t matter
  • Sources: interviews, round-table discussions, literature, observations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q
  1. Develop process maps for each CDVC-segment
A
  • Each process map describes a segment (=process) of the care cycle (=CDVC) and the path that patients follow therein
  • For each step in a process, list all required capacity-supplying resources, both those directly used by patients (“primary resources”) and those needed to make the primary resources available or to support them
  • Process mapping often reveals immediate opportunities for process improvement
  • Sources: interviews, round-table discussions, IT systems, observations
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q
  1. Obtain time estimates for each process step
A
  • Time in minutes each resource spends with a patient at each process step
    o Standard times could be used for common, short and inexpensive steps/activities that vary little across patients, like patient check-ins
    o Actual times should be measured for complex, long and expensive steps/activities, like a major surgical procedure involving multiple professionals
  • Sources: interviews, round-table discussions, IT systems, observations, experts
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q
  1. Estimate the cost of resources involved
A
  • Estimate direct cost per period of all primary resources involved with care delivery
    o E.g. compensation for employees, costs of equipment/supplies
  • Estimate indirect cost of resources required to supply primary resources
    o E.g. costs of supervising, space, furnishings
    o Estimate overhead/support cost of departments/activities supporting primary resources, and assign these costs based on their demands for these departments/activities
    o E.g. central department to sterilize surgical tools
  • Sources: general ledger, (other) IT-systems, interviews
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

6a. Estimate the capacity of each resource

A

Instead of using primary resources’ full or ‘theoretical’ capacity, obtain their practical capacity, i.e. the actual time available for patient-related work over a period (e.g. a month)
* For equipment, estimate the days per month, the hours per day, and then the minutes per month that each piece will be used
* For personnel, obtain the minutes per month by applying the formula below using estimates of:
o total days that each employee actually works each year
o total minutes per day that the employee is available for work
o average minutes per workday for nonpatient-related work

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

6b. Calculate the capacity cost rate (CCR)

A

Reflects the cost per unit of time (usually a minute, see step 4) of the relevant resource

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q
  1. Calculate the total cost of patient care
A
  • Multiply the CCR of each resource (step 6b) by its duration of use in each process step (step 4) to obtain the cost per process step –> leads to
  • Sum the cost of all process steps to obtain the cost per process –> leads to
  • Sum the cost of all processes to obtain the cost for the full cycle of care for a patient with the medical condition
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

Goal of xTDABC

A
  • To gain insight in the costs and cost drivers of the provided care
  • Examine areas of waste reduction and efficiency
How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

Competition and benchmarking

A
  • Competition requires comparison of providers
    o If you don’t have insight in performance, and you can’t compare between providers, competition will not work/ be possible

Benchmarking is method to do so:
Healthcare benchmarking is measuring performance of an organization and comparing it to a standard developed using data from other similar organizations.

The standard can be one specific target or an average

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

Change pathway

A

providers will look at their own performance compared to others and then will be motivated to change their practice. Internal improvement

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
16
Q

Selection pathway

A

insurances will only contract providers with good performances. Automatically, providers with bad performances need to improve to get those contracts and patients as well and “stay alive”.

17
Q

Structure indicators

A
  • Number of beds
  • Stroke unit
  • ED
  • Number of physicians
  • …..
  • …..

Structure indicators are indicators measured once per provider; it is like describing the context of care

18
Q

Process indicators

A
  • Thrombolysis
  • Waiting time on ED
  • CT scan performed
  • Operated within 2 weeks
  • Treated with aspirin
  • …..
  • …..

Process indicators are things that can be measured per patient and they describe the process of care

19
Q

Case mix

A

if you have a more severe patient population and deliver the same quality of care, you still will have worse outcomes as outcomes provider. So, there can be differences between patients in different hospitals that can cause different outcomes. So, for example co-morbidity and a higher average of age causes worse outcomes, even if the quality of care is the same between hospitals.

20
Q

Registration bias

A

when we measure outcomes, we want to make sure that everyone measures the same way and in the same population.

21
Q

How to select patient characteristics

A

1) Predictive of outcome, often disease-specific
2) Differ between hospitals
3) Non-modifiable by hospital

  • Clinical knowlegde and literature
  • Should be (uniformly) measured
  • May differ between outcomes (e.g. quality of life)
  • Adjustment is never 100%!!!
22
Q

Statistical uncertainty

A
  • Statistical method:
  • ‘Random effect’ model = estimates the outcome per hospital, taking into account random variation
  • Small hospitals with an extreme outcome will get an estimate closer to the overall average
23
Q

Conclusion
Outcomes are poor indicators for quality of care

A
  • Crucial to adjust for case-mix and random variation
  • Difficult to improve based on outcomes
24
Q

When we use outcomes for benchmarking: high risk of wrong interpretations

A

Type I error: seeing differences that are not true
Type II error: missing true differences

25
Q

Type I:

A

-Transparency of outcomes, no contract for ‘poor performers’
-Focus improvements on meaningless issues

26
Q

Type II:

A

-Improvements seem to have no effect (demotivating, abandon effective interventions)
-Only differences in costs detected  quality deteriorates

27
Q

Systematic review by Keel et al. (2017)

A
  • TDABC is applicable in health care and can help measure cost in a transparent, flexible and efficient way
  • Applications often constrained by organizational boundaries rather than spanning the full cycle of care
  • TDABC-applications that involve staff are more successful because they allow for tailored implementation and foster ownership & cooperation
28
Q

Principles of cost accounting

A

Cost price = the total costs to produce a product or deliver a service within a company
Consists of cost object(s)

29
Q

cost objects

A

direct costs –> traced to cost objects
indirect costs –> allocated to cost objets

30
Q

Part 1: close reading article Vos et al.
Describe the main aim of this study in your own words.

A

The aim of this study was to quantify the influence of case-mix and random variation on process and outcome measures that are used as quality indicators for breast cancer care in The Netherlands. 3 outcome indicators and 3 process indicators were studied.
Ze kijken naar 6 kwaliteitsindicatoren (=QI = quality indicators). Daar proberen ze de validiteit en betrouwbaarheid berekenen/bepalen. Dan kan je zien hoeveel er echt te maken heeft met quality of care. Je kan dan de kwaliteit vergelijken en daarmee ‘value’ creëren → value based health care. Je wilt dat die QI’s ook echt meten wat ze moeten meten (quality of care).

Case mix = ​ There can be differences between patients in different hospitals that can cause different outcomes. So, for example co-morbidity and a higher average of age causes worse outcomes, even if the quality of care is the same between hospitals.

Random variation = gerelateerd aan statistische onzekerheid. Factoren die dingen die effect kunnen hebben van hoe je op het gemiddelde gaat scoren, van of wel dat je niet weten dat ze er zijn of of weten niet of het effect heeft. We weten wel dat er dingen zijn die random gaan variëren. Daar kan je dan voor varieren. Als je iets bijv heel veel doet worden metingen betrouwbaarder.

31
Q

For each indicator, the rankability is calculated. Explain the term rankability. Should the rankability of an indicator be high or low?

A

Rankability quantifies the remaining “true” between-hospital differences that may be the result of differences in quality of care. It is a percentage expressing the part of heterogeneity between hospitals that is represented by unexplained differences that might be because of the quality of care.
Kwantificeert de resterende “echte” verschillen tussen ziekenhuizen die het gevolg kunnen zijn van verschillen in kwaliteit van zorg. Het is een percentage dat het deel van de heterogeniteit tussen ziekenhuizen uitdrukt dat bestaat uit onverklaarde verschillen die het gevolg kunnen zijn van de kwaliteit van de zorg.
Als dit hoog is zijn de vergelijkingen betekenisvoller.
Van de 100% verschillen tussen ziekenhuizen, hoeveel % is echt een verschil tussen quality of care tussen ziekenhuizen? Dit % moet dus zo hoog mogelijk zijn. Betrouwbaarheid van het rangschikken.
Rankability = richt zich op de betrouwbaarheid van het rangschikken.
A low rankability means case-mix and random variation explain the majority of hospital differences and a minority could possibly be explained by quality of care. It can be concluded that such an indicator tells us very little about the quality of care and is not reliable (ie, a large random variation) or valid (ie, a large case-mix effect).
A high rankability also does not necessarily represent true hospital differences because it may include residual confounding such as unmeasured case-mix (eg, comorbidity) or other unknown differences between hospitals.
Rankability of an indicator should be high, because it suggests that a greater part of the differences are explained by actual differences.
Rankability also increases by increasing the number of events. To illustrate this, the number of events per hospital strongly influenced the reliability of the QI 1 score. Rankability also increases by increasing between-hospital variation. Because the incidence of irradicality after BCS is low, there is little between-hospital variation. The number of events and the between-hospital variation can be increased by evaluating multiple years of data.
Om het pure effect van QI’s te meten, moet je adjusten voor case-mix en random variation.
Random variatie gaat over een kansspel. Je kan hiervoor adjusten/corrigeren door je datasample te vergroten.
Random variation prevents an indicator from producing the same result on repeated measurements, thereby making the indicator less reliable.