Term
| What are the potential advantages of Computer Decision Support Systems (CDSS)? |
|
Definition
- increase patient safety
- remove variations in service
- increase quality of care
- to make the relationship between cost and benefit more clear
- helps relieve financial cost through efficiency |
|
|
Term
| How does a "before and after design" observational study evaluate the CDSS and what are the problems associated with this type of study? |
|
Definition
- "Before and After Design" study compares the outcomes before and after implementation of the CDSS.
- Results can be misleading because trends and changes of patient care and physician performance occur over time. |
|
|
Term
| How is a "time series design" observational study carried out in relation to a CDSS? |
|
Definition
A CDSS acts as the intervention and it is applied intermittently over a period of time alternating with control periods of time.
This helps correct the problem of trends changing over a shorter amount of time, but it does not eliminate the error created over a longer period of time. |
|
|
Term
| When physicians are separated into groups that utilize a CDSS and groups that do not, they are being observed in what type of study design? |
|
Definition
Clustered randomized control trial
|
|
|
Term
| What are difficulties when carrying out a clustered randomized control trial? And what can be done to remedy the complications of those difficulties? |
|
Definition
- difficult to get enough groups together to have a large enough sample size so that prognostic factors are balanced
- A multivariate analysis can be done to account for the differences in practice methods |
|
|
Term
| Are patients or physicians the desired unit of analysis in a Clustered RTC? |
|
Definition
Physicians
- because if you use patients as the desired unit of analysis you are essentially evaluating the physician and not the CDSS |
|
|
Term
| A common problem with CDSS's are that physicians stop using them during an observational study. What analysis can be done to remedy this problem? |
|
Definition
Intention To Treat (ITT)
- because it provides equal sample sizes and prevents overestimation (preserves balance of groups) |
|
|
Term
| Why are contamination and co-interventions a problem with Randomized Control Trials that evaluate CDSS? |
|
Definition
Since physicians are the target of the intervention rather than patients, they are not blinded to the intervention and can influence the results of the study. |
|
|
Term
| A physician sees thirty patients in marathon 14-hour day at the clinic. Half of the patients were a part of the control group and did not have alerts on their electronic health record. The other half did. However, the physician remembered a previous patient with an alert and treated a control patient differently because of the previous alert. What bias is the physician illustrating in relation to an RTC evalutation of a CDSS? |
|
Definition
|
|
Term
| How would a researcher solve the problem of contamination (bias)? |
|
Definition
|
|
Term
| How does co-intervention create a bias during an RCT for the experimental group? |
|
Definition
Physicians who receive alerts may do more than just the standard treatment because of the CDSS reminder in comparison to the standard level of care an aware and careful un-reminded (control group) physician might give. |
|
|
Term
| How can a researcher remedy co-interventions as a bias during an RTC evaluating CDSS? |
|
Definition
Tell physicians possible co-interventions and what is allowed ahead of time so both the control and experimental groups are aware of possible options for treatment.
OR
Blind everyone. (physicians, outcome raters, administrators) |
|
|
Term
| What is the bias called when the patient file is more complete for the experimental group than the control group during an RTC evaluating CDSS? |
|
Definition
|
|
Term
Which of the following describes the bias that occurs when the experimental group is checked more often/thoroughly for the desired outcome?
A. Contamination
B. Co-interventions
C. Data completness bias
D. Surveillance bias |
|
Definition
|
|
Term
| Process of care outcomes are measures of physician compliance with health care processes (did the doctor do what they were supposed to do?). What are some examples of standard health care processes a primary doctor should do? |
|
Definition
-follow set immunization rates
- adequate cancer (breast, prostate) screening rates
- patient education/counseling to reduce further risk |
|
|
Term
| A CDSS was developed at Princeton Plainsboro Hospital to help regulate and provide guidelines for the administration of vicodin during pain-management treatment. However, when this CDSS was transferred to New York General Hospital, the CDSS failed to successfully increase the level of care as it did previously. What did this CDSS lack? |
|
Definition
Exportability- CDSS must be easily integrated into existing software at a new site, use established automated databases, be maintained by people at the new site, and have physician acceptance |
|
|
Term
| A newly implemented CDSS improved practitioner performance by 57% in the Catholic Healthcare West system. Would the improved patient outcome be expected to be higher or lower than the practitioner performance increase? |
|
Definition
It is expected to be lower than the practitioner performance increase because CDSS's are beneficial mostly in reminder and drug dosing systems rather than diagnostic systems. Therefore, physician performance (in the easily forgotten/mistaken areas) would improve more than patient diagnosis and treatment. |
|
|
Term
| What are some of the barriers to the spread of CDSS in our increasingly technologically dependent world? |
|
Definition
- "bottom up" apporach (hospitals have to design and implement their own systems which are then supposed to be connected to a national health information network)
- Cost
- Privacy/Confidentially (HIPAA)
- Implementation
- Lack of workforce initiative in implementation
- Lack of incentive |
|
|
