Perinatal Collaborative July Coaching Call July 27 th , 2016 - PowerPoint PPT Presentation

LHAREF HEN Perinatal Collaborative July Coaching Call July 27 th , 2016

Agenda for Today’s Call Welcome and introductions Update on Our Data: Taking action in light of variation & reliability Emerging Issues Office Hours (last 30 minutes) – Please stay on the line to share what you are working on, learn from other teams, and ask questions

Welcome and Introductions Welcome from the IHI and LHA team Who’s on the call? – Please chat in your name! Emerging Issues?

Update on Our Data: Taking action in light of variation & reliability

Variation Reliability

Common Cause • Common cause variations are problems built right into the system, such as defects, errors, mistakes, waste and rework. In a stable system, common cause variation will be predictable within certain limits • Common cause variation is produced by the aggregate of the variation in all the variables VARIATI VA IATION ON (random) exi xists sts Special Cause • Special cause variations represent a unique event that is outside the system, such as a natural disaster, or an unexpected strike by public transportation workers • Special cause is produced by a non-typical variable. (non-random) Knowledge: Improvement Action 1. Random variation: Change the system 2. Non-random variation: Investigate

1. Most of the time in improvement we are trying to introduce non-random variation into the system towards the direction of goodness. And then making it stable (random) at a new level or with less spread.

RUN CHART TOOL: Three (3) simple rules that indicate if something is not typical random variation. Only one rule needs to be present. Rule 2 Rule 1 Median Median Shift : the purpose of this test is to Trend: The purpose of this test is to identify a shift in the process. A run identify a low-probability trend in the data containing 6 or more consecutive data set. A trend is defined as 5 or more points all above or all below the median consecutive points constantly indicates a non-random pattern in your increasing or 5 or more consecutive data which should be investigated. This points decreasing. If a trend is detected it non-random pattern may be a signal of indicates a non-random pattern in your improvement or of process degradation. data which should be investigated. (The IHI extranet) (The IHI extranet definition) Rule 3 Median Astronomical Point(s) An astronomical point is one that is obviously and blatantly much higher or lower than all the other points on the chart. On a run chart this rule is not determined statistically but rather by judgment or consensus. (The IHI extranet definition)

Robust Testing and Reliability Tiers Full, Sustained Implementation Redesign from Failure A P Modes > 99% S D Identify critical failures & then redesign Later tests designed to Predict and prevent failures Identify Failure and Mitigate (90 – 95%) Human factor changes A P Redundancy function Test over a variety of S D conditions to identify weaknesses Prevent Initial Failure (80 – 90%) Early tests are simple Using intent and standardization Designed to learn & succeed Segmentation Hunches Theories Ideas @ V Crowe

Prevent Initial Failure CRUD • C omplex • R edundant (stupid Unnatural redundancy) • U seless Variation • D elays @ V Crowe

Prevent Initial Failure (80 – 90%) Associated Changes • Using intent and • Common equipment • Standard orders standardization • Standard rooms • Segmentation • Memory aids • Personal check lists • Awareness • Training • Compliance Feedback @ V Crowe Reference: IHI Reliability Presentation

intent and standardization Necessary not sufficient Factors Affecting Human Vigilance : • Reliance on memory • Distractions / interruptions • Fatigue • Sleep deprivation • Shift work • Lack of training and experience • Overload • Psychosocial factors Reference: IHI Reliability Presentation @ V Crowe

Identify Failure and Mitigate 90-95% Associated Changes • Human Factor Changes • Decision aids & reminders built into • Redundancy Function the system • Differentiation e.g. color coding • Constraints and Affordances • Desired action the default • Alarms • Provide clear visual or other sensory clues • Schedule key tasks • Take advantage of habits and patterns • Continue standardization of processes • Create intentional redundancy (carefully) and process for mitigation @ V Crowe Reference: IHI Reliability Presentation

“This is a mop sink www.mistakeproofing.com @ V Crowe

unknown Patients should experience healthcare processes that are more reliable than manufacturing processes. www.mistakeproofing.com @ V Crowe

Redesign from Associated Changes Failure Modes > 99% • There is a process to detect failure and Identify critical measure the failure. failures and then • There is a clearly articulated process by which the knowledge about the defects redesign get back into the design • FMEA – one method • Failure modes • What could go wrong? • Failure causes • Why would the failure happen? • Failure effects • What would be the consequences of each failure?) @ V Crowe Reference: IHI Reliability Presentation

Redesign from Failures @ V Crowe

Update on Our Data: Not everything that can be counted counts, and not everything that counts can be counted. Albert Einstein, Physicist (attributed)

PC.01

Time Period min max average_team median n 10 - 2015 0 14.29 0.98 0 30 11-2015 0 20.0 1.78 0 29 12-2015 0 28.57 2.29 0 29 1 - 2016 0 16.67 0.82 0 30 2 - 2016 0 4.44 0.29 0 29 3 - 2016 0 9.09 0.52 0 29 4 - 2016 0 100 5.2 0 25 5 - 2016 0 50.0 4.56 0 19 6 - 2016 0 0 0 0 6

Hemorrhage

Time Period min max avg. median n 10-2015 0 41.67 2.87 0 24 11-2015 0 9.01 1.48 0 24 12-2015 0 13.79 1.42 0 24 1 - 2016 0 10.64 1.38 0 24 2 - 2016 0 13.14 1.35 0 22 3 - 2016 0 19.61 2.04 0 22 4 - 2016 0 13.70 1.68 0 20 5 - 2016 0 18.69 2.42 0 16 6 - 2016 0 11.24 2.25 0 5

HEN OB Hemorrhage Zero One or more events reported events reported Hospitals with Hospitals with population population per month per month (D) less than Population(D) (D) less than 100 patients 100 patients and had one < 100 month and had zero events or more events reported. reported.) Hospitals with Hospitals with population Population (D) population per month per month (D) greater > 100 month (D) greater than 100 than 100 patients and patients and had zero had one or more events events reported reported

HEN OB Hemorrhage Zero One or more events events reported reported (42% ) (58%) 4 8 Population(D) < 100 month (3 had 1 event; ( 4 < 50 and 4 > 50) 1 had 2 events) Population varied from 20 per month to 90+) 3 11 Population (D) > 100 month ( 3> 200) (3 > 200 and 1 > 600) (1 had only 1 data point)

Pre-eclampsia

Time Period min max Avg. median n 10 - 2016 0 166.67 13.44 0 23 11 - 2016 0 34.48 1.64 0 21 12 - 2016 0 136.36 6.49 0 21 1 - 2016 0 142.86 19.53 0 22 2 - 2016 0 1000 54.11 0 21 3 - 2016 0 500 51.69 0 19 4 - 2016 0 86.96 5.44 0 16 5 - 2016 0 250.0 32.38 0 15 6 - 2016 0 0 0 0 4

Zero One events or more events reported reported 54% 46% Population (D) 8 5 in single digits (0- 9) HEN (1 or 2 events) Pre- Population (D) 2 3 ecla lampsia in single/double digits (1 or 2 events) (0-20) 2 4 Population (D) in double 1 – 3 events pop 17 – 36 (10 - 50) digits 19 events pop. 22 – 55

OB Trauma with Instrument

Avg. Time Period min max median n team 10 - 2015 0 400 48.05 0 24 11 - 2015 0 500 82.77 0 25 12 - 2015 0 333.33 66.62 0 27 1 - 2016 0 1000 95.70 0 27 2 - 2016 0 250.00 37.40 0 25 3 - 2016 0 333.33 26.29 0 23 4 - 2016 0 1000 76.16 0 22 5 - 2016 0 428.57 39.92 0 17 6 - 2016 0 333.33 68.03 0 7

Zero One events or more events reported reported 24% 76% Population (D) 7 11 in single digits HEN (1 – 8 events) (0 – 9) OB Trauma Population (D) 7 0 wit ith in single/double digits (1 - 6 events) In Instrument (1 – 25) 0 4 Population (D) in double (1 – 4 events) Digits (10 – 38)

OB Trauma without instrument

Time Period min max Avg. median n 10 - 201 0 700 38.50 1.24 30 11 - 2015 0 1000 42.96 0 30 12 - 2015 0 769.23 31.15 0 30 1 - 2016 0 840.00 40.70 0 29 2 - 2016 0 944.44 47.02 0 29 3 - 2016 0 800.0 39.73 0 28 4 - 2016 0 55.56 10.08 0 25 5 - 2016 0 166.67 14.99 0 19 6 - 2016 0 34.78 7.01 0 7

Zero One events or more events reported reported 10% 90% Population (D) 3 6 in single digits HEN (1 – 8 events) (0 – 9) OB Trauma Population (D) 14 0 wit ithout in single/double digits (2 – 15 (101)) In Instrument (1 – 25) events) 0 7 Population (D) in double (1 – 22 events) Digits (10 – 38)

Summary Thoughts We are seeing things we did not see ... We have improved our clinical processes ... We have much we can learn from each other ....

Emerging Issues Pre-eclampsia categories ...