IHI.org - A resource from the Institute for Healthcare Improvement
Header Image






Testing Changes

Keys to successful implementation of protocols or care standards include the following:

 

  • When possible, base them on firm evidence from the scientific literature.
  • Design them using a multidisciplinary team.
  • Involve all stakeholders, using an information feedback process to facilitate everyone's trust, confidence, and buy-in.
  • Test the protocol using small plan-do-study-act (PDSA) cycles in the clinical environment, modifying the protocol as needed to make it unambiguous, safe, and acceptable to practitioners.
  • Measure the impact on work processes and outcomes, and feed the information back to the users.
  • Test vigorously using multiple small tests, in order to reduce problems with safety and compliance to a minimum before full implementation.

 


Using the Plan-Do-Study-Act (PDSA) Cycle

Example: Use an insulin drip protocol to achieve appropriate glucose control in hyperglycemic patients.

 

Plan
Use an insulin drip protocol to control the glucose in a hyperglycemic septic patient, keeping it less than 180 mg/dl as measured at the bedside.

Do
Use the protocol on one patient, recording insulin drip changes, amount and timing of insulin boluses, and blood glucose levels and timing. Also, document whenever the protocol was not followed and the time and reason for the deviation from the protocol.

Study

  • The protocol resulted in two incidents of mild hypoglycemia.
  • After the episodes of hypoglycemia, the nurse violated the protocol when the glucose was dropping rapidly by reducing the incremental drip change and omitting the bolus to avoid this complication.
  • The protocol was ambiguous in a few places that the nurse felt could lead to mistakes.
  • The protocol was very effective at reducing the glucose to below the 180 mg/dl threshold.

Act

  • After discussing findings, the team modified the protocol to adjust less aggressively.
  • The team removed areas of ambiguity in the protocol.
  • The team is ready to plan the next small test of change.

 

Example: Implement a process to maintain blood glucose of less than 180 mg/dl consistently in hyperglycemic patients with severe sepsis and septic shock.


Cycle 1:
Establish a system to monitor and document blood glucose measurements in critically ill patients.

Cycle 2:
Modify an insulin drip protocol, perhaps adopted from a colleague, to maintain glucose concentrations in hyperglycemic patients to less than 180 mg/dl and send it out for comment and buy-in to practitioners who will use the protocol.

Cycle 3:
Test the protocol on one or two patients and modify as needed to improve safety and address objections to workflow problems.

Cycle 4:
Because of problems with frequent glucometer checks and difficulties in access to the instruments, purchase more for the unit.

Cycle 5:
Because there is overshoot in glucose levels when 50 percent dextrose is used to treat hypoglycemia in the test patients, modify the protocol to address this problem and measure the results.

Cycle 6:
Continue small tests and modifications until safety and consistency are established, then release for general use.