Puneet Vohra

When AI removes one constraint ,it often creates another: Taking the example of US healthcare insurance process , lets us understand the reasoning of this point Reason - Artificial intelligence can streamline claims processing eligibility checks, reducing human error and speeding up approvals. But these incorporates new obstacles such as algorithmic bias ,data privacy concerns and dependency of technology infrastructure Example - AI based solutions can eliminate delay in taking prior authorization caused by manual reviews. But this solution will face constraint If y if there are missing necessary details by the Claimant or non -standard data are technically unfairly denied coverage because the algorithm cannot interpret nuanced medical conditions Solution - So when AI removes operational bottlenecks, it parallelly throws the constraints such as fairness, transparency and compliance. Additionally, US health care process Insurance companies must implement explainable AI models, conduct Bias Audits, ensure compliance with HIPPA and CMS guidelines. Human in the loop or we can human oversight should complement aid decisions to maintain fairness. Thanks

So I am taking the example of Health care domain: Background: In Health insurance companies the claim processing systems often denies certain Medical procedures due to missing or unclear prior authorization details. These denials triggers the Manual Appeals process, where the Appeal resolving analyst handles review and either uphold or overturn the denials based upon the additional medical documents submitted by providers or members. Repeated Human exceptions: Analysts notice that a particular type of CPT code for spinal injections (ex -99213) was being frequently denied due to missing prior authorization, , but during the Appeals : The provider is categorically indicating that is a urgent Need and submit in the clinical notes as well Authorization had been requested but delayed The procedure was performed in an emergency setting, exempting it from pre-authorization. Impact - More than 80 % of the cases were reversed the denial on Appeal after reviewing the additional notes/data. Human in the loop learning loop: Initial AI performance The ML/AI based machine will flag the appeals with missing Prior authorizations and valid denials, offering no recommendation for reversal Human interventions to modify the AI generated outcomes: Analysts consistently Overturned(Approved) those cases which contains structured Comments such as mentioned below: " Emergency setting, Prior Auth not required policy X" "Provide er submitted auth request but payer systems delayed it" Pattern Identification: The AI system uses NLP(natural language processing) to read clinical notes and Appeal attachments/documents. Its resembles the keywords such as Clinical urgency phrases, emergency settings , or Auth request timestamps with Overturned decisions Model Re-training: AI model is required to retrained periodically with these type of examples which eventually results in: Achieve the probabilities of that a denial will be overturned Suggest to Auto escalate for Appeal or even can Pre-approve under certain conditions Impact: Reduction in unnecessary manual Appeals by 30 % Min. Faster resolution times for Providers and patients Improved trust from Providers due to fewer incorrect denials

Yes! While writing the problem definition in Define phase of DMAIC. There can be different opinions of different stakeholders. Here are some practical steps which can be utilized to avoid initiating a project with misunderstood problem: Root Cause Analysis: Use Techniques such as 5 Whys or Fishbone Diagram to identify the root causes of the issues by ensuring that each stakeholders actively participate in this analysis to provide their input Data Collection: Gather the volume number and defects data as well to understand the problem with the lens of data Facilitate workshops: Encourage the people to speak their opinions on the problem and group their thoughts based upon similarity Problem statement Drafting: Draft a problem statement based on the insights gathered Ensure that problem statement is Specific , measurable achievable relevant and timebound Validation and Consent buying from stakeholders: Present the drafted problem statement to all stakeholders for validation Encourage feedback and make necessary adjustments to ensure everyone agrees on the problem definition Documentation: Document the agreed-upon problem statement and ensure it is accessible to all stakeholders Include the logic and supporting documents of problem definition To avoid misunderstood Problem: Do a transparent and open communication throughout the Define Phase Regular Check ins Be prepared to refine the problem statement iteratively based upon the new insights or feedback

Gantt Charts provides a visual timeline, it makes easy to see the sequence of activities, the durations and dependencies. Its helps in breaking down the project in to manageable tasks, assigning responsibilities and setting deadlines. When we update the chart project managers can monitor progress and recognize delays and make necessary adjustments It serves as communication tool helping stakeholders to understand the project status. How can we overcome the limitation of Gantt chart, while we are working on complex dynamic projects : We have to ensure the Gantt chart is continuously updated to reflect the current status of the project Use Gantt charts in conjunction with other project management tools like Kanban boards, task lists and Agile methodologies to provide a more comprehensive view Be ready to change the Gantt charts as the projects evolves. This reevaluating task durations, dependencies and resource allocation. Incorporate risk management practices to predict and resolve challenges that are impacting the project timeline Regularly communicating with stakeholders to ensure they are aware of changes and can provide support as required focusing on key milestones rather than indulging in the details of every task. This helps is maintaining the high- level view of the project's progress.

Type of Analytics Description Role in Decision Intelligence Example in U.S. Health Insurance & Medical Billing Descriptive Analytics Analyses historical data to understand past events. Provides context but lacks actionable insights. Tracks claims, appeals and volume trends, customer demographics, and common claim rejection reasons. Helps insurers identify historical patterns, but doesn’t directly drive decisions. Diagnostic Analytics Explores reasons behind past outcomes. Helps identify root causes but is limited in proactive guidance. Analyses reasons for claim rejections (e.g., missing data, billing code errors, Medical necessity, Out of network hospitals, Experimental treatment, Coverage exclusions) to reduce denial rates. Gives insights into process issues but doesn’t offer action plans. Predictive Analytics Uses historical data to forecast future. Helps in anticipating future outcomes but lacks recommendations on actions. Predicts claims and Appeals likely to be denied, allowing proactive error-checking. Forecasts high-cost claims and Performance Guarantee fixed TAT for resource allocation, but doesn't directly suggest how to handle the predictions. Prescriptive Analytics Recommends actions based on predictive insights. Directly helps in Decision Intelligence by providing actionable steps for desired outcomes. Recommends policy/process improvements to reduce denial rates, such as automated error-checking for commonly rejected claims. Suggests optimal billing codes to reduce processing time and improve revenue cycle management. Decision Intelligence in Action Example: In health insurance domain , a DI framework might combine predictive analytics (identifying high-risk claims and appeals, specific scenarios state specific) with prescriptive analytics (suggesting error-checking steps) to enhance decision-making by reducing claim delays and improving accuracy. Overall, Prescriptive Analytics plays the most crucial role in DI, as it not only anticipates future scenarios but also provides actionable guidance which is programmed from Machine learning, directly impacting the quality and speed of decision-making.

There are the limitations in today’s LLM explaining in the context of Health insurance and medical billing: Context Understanding of Complex Insurance Claims and Appeals: LLMs cannot find the specific case details over extended conversations or documentation review. This may lead to errors in responses about a patient's claim status or coverage details after long interactions. Complex Reasoning in Claims Processing and Appeals: LLMs struggle with multi-step reasoning required for processing complex claims or appeals. Example: It can fail to accurately check the eligibility for coverage based on nuanced policy clauses or miss crucial steps in appeals cases. Ambiguity and Outdated Information on Health Policies: Limited in handling updates on recent policy changes, which are frequent in healthcare. Example: May provide outdated information about coverage for new treatments or evolving healthcare guidelines (e.g., telemedicine coverage changes during the pandemic). Ethical Considerations and Patient Privacy: LLMs lack intrinsic ethical understanding and may inadvertently mishandle sensitive patient data or protected health information. Risks include recommending actions that may conflict with HIPAA guidelines or fail to consider patient privacy concerns. Novelty and Out-of-Scope Knowledge in Medical Coding and Procedures: LLMs cannot independently interpret complex medical codes or new procedures. Example: Limited in interpreting new CPT codes or recent healthcare treatments, affecting accurate claim submissions and billing.

Giving you an example from Health insurance industry particularly in claims processing and Appeals handling processes: Claims Processing: Reducing Delays and Enhancing Accuracy When Applying the Systems Thinking: The claims processes involves various stages such as Claim Submission by physician or patient electronically , then the Adjudication also happens in Adjudication engines of the insurance companies. System thinking identifies interdependencies and potential bottlenecks such as lengthy manual reviews by the clearing houses, not clear claim submission process steps , incomplete patient data. By Mapping the end to end claims process, insurers can find recurring issues that results in delayed and inaccurate submitted claims also as inadequate training for clams adjusters or technology drawbacks When Applying Design Thinking: After Understanding systematic issues, design thinking can bring customer and employee feedback to the forefront, identifying pain points for both policyholders and claim adjusters. Through ideation and prototyping insurers might test AI powered tools for automated claims and Appeals intent identifying and validation or enhanced user portals that guide providers on correct submission practices, ensuring completeness. Testing these solutions ensures they are user-friendly and effectively address core needs Outcome: Streamlining claims processing reduces processing time of appeals and claims. It reduces error rates but also boosts customer satisfaction as claims are handled accurately and quickly Appeals Handling: Improving Transparency and reducing rework Applying Systems Thinking: Appeals handling are maximum number of times found complex due to regulatory requirements and varied medical necessity criteria. Systems Thinking helps in end to end mapping of Appeals journey in the entire Appeals ecosystem, considering the roles of claims adjusters, medical clinical reviewers, business partners and regulatory compliance officers. This analysis reveals where communication breakdowns or repetitive rework (e.g. Multiple review stages for the same appeal) slow down the process and frustrate both providers and Members. Appeals Handling Using Design Thinking: Once bottlenecks are identified, Design Thinking introduces human- centred solutions by engaging stakeholders- providers, members and Appeals resolving analysts. Outcome: A transparent efficient Appeals process reduces frustration and improves provider relationships and lowers administrative costs by cutting down on repetitive work. Members/ feel satisfied as they transparent updates and providers can more effectively advocate for their patients. Improving Member Engagement with preventive care incentives Applying systems thinking: Preventive care initiatives in health insurance involve numerous factors from care accessibility to customer satisfaction. Systems thinking can help insurers how preventive care programs interact with other insurance functions like customer communication, provider networks and data analytics. An Analysis might show that low engagement is linked to a lack of awareness, ineffective communication channels or insufficient data to target specific at risk populations. Applying Design Thinking: With these insights, insurers can use design thinking to empathize with Members, understanding their barriers to preventive care engagement. By brainstorming and testing solutions like personalized health reminders wellness programs or incentives for preventive screenings insurers ensure these initiatives resonate with Member's needs for example sending test reminders or offering reduced co-pays for preventive visits might be piloted and refined based on member feedback Outcome: A higher engagement rate in preventive care programs leads to healthier members, reducing long term claim costs and improving member satisfaction with the insurer.

The below answer is portrayed considering the 'Health insurance and medical billing' process to draw a comparison between standard work and work instructions: Aspect Standard Work Work Instructions Definition The group of activities illustrating the best way to handle claim submission, medical billing and denial management It comprises the detailed explanation of how to perform each and every process/step within the medical billing and insurance process Scope Broadly it covers steps such as claim submission, coding, insurance verification and payment posting It covers detailed steps such as claims form filling, claim uploading in portal, claim receiving in clearing house, claim processing and adjudication, payment posting , letter sending Purpose The purpose of standard work is to standardize the process for all teams ensuring consistent handling of claims, appeals and other workflow activities It provides the guidance in terms of specific process steps and compliance guidelines too with payer rules Level of detail L1 level steps emphasizes on activities such as verifying insurance, coding, submitting and reviewing denials and submitting them to payers etc. L4 and L5 level which provides click level information of each process and also includes the screenshots and images to facilitate the reader. Example Workflow for claim submission including verifying insurance, claim adjudication, Claim error verification, coding submission by medical biller of doctor and reviewing denials Instructions of coding claims such as ICD - 10 and CPT codes, it includes specific payer rules and software usage guidance also Purpose in process Standard work is a practise to ensure the process which is free of error and accelerating the claims processing It ensures the accurate execution of specific tasks such claim form filling , claim verification, policy coverage verification, claim validation and payment processing and letter generation Audience To get the high-level view of the process for same set of practices to be followed by all team members The team which follows the work instruction need to know the payer rules, application and software use for case handling

Conformity Bias: It's a group activity done by the team to align on everyone's ideas and thoughts often preferred to follow the critical thinking and personal judgement. While doing the project management, conformity bias impacts the decision making in several negative ways: 1. Creative ideas and solutions which can transform the entire business are not prioritized. 2. Conformity creates the random chances of project failures, it creates lack of ownership 3. Conformity bias directs team to follow the single solution by ignoring the cons or any potential risks around it. 4. It narrows down the chances of doing any study by disaggregating the process and fundamentally asses each task type is either of follow-up, complex research, rule based and predictive Strategies for project managers to mitigate conformity bias: Once the decision making is done its mandate to review the decision whether Conformity played any role. What went wrong? what went good? In order to design the future decision with takeaways and calculated risks. Always prefer forming teams with different backgrounds and perspective in order to view from different viewpoints. The diversity in team formation bring reduces the conformity as the different perspectives avoid the reduction of dominant or single thought or idea always. During brainstorming any problem always prefer team to provide their own idea or way out to come to the solution as this avoids conformity bias and a good boost to initiate the brainstorming Keep taking the anonymous feedback from the team here you are either providing a new product or service, that helps to confirm their individual preference rather being a follower Form a strategy where in you dedicatedly assign someone from team to challenge the views to ensure that diverse perspectives are implemented Build a culture where open question from anyone is appreciated Conclusion: If we follow the above listed strategies which concludes that individual idea is heard then success rate of any project gets high and we can see even organizations such as 'Google' ,'NASA' ,'Microsoft' and 'Open AI' etc. urges their hiring teams to have the diversified resources.

'Workflow Analysis plays a crucial role in shaping the DMADV framework: Define phase: Work flow analysis helps to understand closely the As is flow of the business process. Example: If a health insurance company is going to develop a super-fast claim processing and resolution system that has been analyzed by seeing the As Is flow that claims resolutions are often delayed Then the problem definition in business case would be reducing processing time by 20%. Measure phase: If the workflow analysis is done during this phase that helps in collecting data of actual business metrics, also this data is utilized as benchmark for the new process Example: If the claim resolution of treatment and hospitalization expenses is taking 30 days. This is our Measurement baselining. The Core team working on projects can aim to reduce the claim resolution time by 50 %. Analyze Phase: In Analyze phase, workflow analysis is done to recognize the waste of entire process from the lens of 'People, Process, Technology and Control' Example: An analysis might show that manual claim form adjudication or processing is a bottle neck process. This insight can lead team to design of an automated claim form processing by leveraging the OCR and AI to create the automated data entry which will impact the efficiency of the process. Design phase: In design phase the outcome of workflow analysis can help us to re-write the 'To be flow' and develop the solution considering the recognized gaps and needs. Example: If the hospital or patient want that they should get the real time update of claim processing as well then the solution which is designed must have the automated notification sending feature incorporated. Verify Phase: The workflow analysis helps to test the designed solution as per the Business requirement , its makes sure the 'To be process' meets customer expectations. Example: Once the new claim processing automated solution is 'Go live'. The core members can utilize the new process reduces the resolution time and improves the customer satisfaction of the claimant. Conclusion: By integrating the workflow analysis into the DMADV framework, the risks are minimized and increases the probability successful deployment of solution in operations to resolve and process the claims of expenses done by the hospital and patients which ultimately reduces turn around time.

Tap root Analysis - Tap root analysis and Root cause analysis is the same thing used when we work on any issue/problem. The ultimate moto is to find the root cause of the problem. It is used very often by quality and process excellence professionals to improve the business. In this analysis the problem is seen from each lens i.e. People, Process, Control and Technology and then under each lens the points are listed. Once this is done then the classification of all causes is done in to the 'Data door' and 'Process Door' Steps: Recognize, understand and define the problem. Collect the relevant data and details pertaining to that recognized issue. Make plan which comprises of sequence of activities leading up to the problem this helps in understanding where things went wrong. Develop a chart that visually maps the various factors that contributed to the incident. Make use of Fishbone diagram, Why-Why Analysis, Fault Tree Analysis to dig deeper in to the reasons behind the failure To offer and develop the solutions or actions which eliminates the root cause ensuring these actions are feasible and sustainable. To sustain the implemented solution by continuously monitoring initially for 1- 2 months ensure that problem does not recur. Benefits and limitations of Tap root analysis, Root cause and Why-Why Analysis: Benefits Limitations Root Cause Analysis Ideal for team brainstorming sessions to dig the problem For Manufacturing Industry - Causes are classified in Man, Machine , Material , Method For Service Industry - Causes are classified in to the People, Process, Control, Technology Encourages the collaboration with teams of upstream and downstream process Identifies probable causes however does not dig in to the each one. Does not provide the sequence of activities leading to the identified problem Tap Root Analysis Systematic approach for comprehensive analysis Provides multiple causes Easy representation of causes through charts of complex problems Prevention oriented Complex issues and highly technical problems Due to its depth takes longer time Complex required specific training to execute the analysis Why-Why Analysis No specialized training, very simple approach Provides only one root cause. Can be done very fast for smaller problems Very much subjective Focuses on cause and effect relationship by deliberately asking 'why' Not favourable for highly complex issues Tap Root analysis are better suited for below mentioned cases: Cyber security breach in Finance domain Safety Bypass in Manufacturing industry Surgical errors in Healthcare domain Hose failure in High speed racing cars Aviation Safety issues in aircrafts Space crafts Helium gas leakage Railway accidents Gas leakage from the chemical processing plant

Control charts are are charts which are shown with average, upper and lower control limits/lines to see whether process is stable or not. Generally, the control charts are X-bar, R- bar used for continuous data such as measurements, weights and time. P - Chart and C- Chart used with discrete data such as Fail/Pass cases, number of customer complaints Point-Using Control Chart is not mandatory for every Lean six sigma Gb or BB project in health insurance claim processing, but they highly beneficial and recommended. Reason- Control Charts help in getting the process stability, data driven decision making, and recognize variations that required immediate corrective actions. But their need depends upon the specific goals and context of the project. Example: To reduce the time taken to process health insurance claim, Control charts can be a useful tool. It can easily check the average processing time and recognize the waiting time and variations in the process. If we get the output of control chart in acceptable limits then our process is stable. Or If the points are outside the control limits, then this is the special cause variations and that needs to be addressed. In contrast to this, If there is a small project which focuses in improving the interactions of customer service, then use of control charts might not be that critical. Moreover when the process is not involved in continuous monitoring. In summary, Since control charts are a impactful tool in Lean Six Sigma, but their use should be checked based on the specific needs and context of the project .

ADAKAR Model:- This model was invented by 'Prosci'. It manages the smooth transition of the process from one platform to another ensuring that everyone is prepared and supportive. A - Awareness of necessity for the change D - Desire to contribute and support the change K - Knowledge that how can we change the process A - Ability to implement the skills with good behavior R - Reinforce the sustenance system that monitors the process and do the fine tuning also Integrating the ADKAR Model with DMAIC methodology enhance the effectiveness of change management: D- In Define phase the project leader should bring in notice to core team members that by inculcating awareness for need to change by very clearly articulating the problem and its tangible and intangible impact. M- In Measure phase the we have to drive the hunger of desire by engaging all stakeholders and measuring their readiness and willingness to change. Because when we highlight the benefits of the project that gives an encouragement to team A- In Analyze phase, we should share the knowledge with team by analyzing the data and giving insights on change that how it will be impacting the process. We have bring that conviction in everyone mind with proper rationale by doing the root cause analysis Improve: Develop the team's ability by giving them required training and giving them necessary skills to think and build effective solutions using best technology Control: Build a system which is reinforced with proper controls and feedback mechanisms to sustain the change In the conclusion I would like to state that if we integrate the 'ADKAR' with 'DMAIC" we can actually involve people who are ready, willing and able to embrace the changes that leads to more successful and sustainable outcome

Gamification: Gamification is doing short duration games or exercises by distributing participants in teams and giving them a exercise to apply the theory concepts in practical relevance . It enhances the learning to the next level and also a platform to participants to open and clear their doubts with practical exercises. Several unique ways which can be applied for Gamification: Giving small quizzes during training so that participants can critically think over the important points with logical reasoning Creating an environment of messy process on any platform where participants can see and identify the different wastes of lean six sigma Asking Lean six sigma based questions on portal and rewarding something which inspire them to continue lean and answering these questions To promote games on fishbone analysis, control charts and process mapping , as these are pretty simple and complex problem solving tools Provide recognition such as DMAIC, Value Stream Mapping, or FMEA to inspire learners to learn more and grow faster Share a process problem, and participants must solve within the speculated time to find the root cause using Lean Six Sigma tools like Pareto Charts and 5 Whys. Rewards should be based on speed and accuracy. To use dashboard or progress bars/indicator of all participants, to encourage the contribution in leaning Lean Six Sigma practices Different roles can be given such as Project leader, Project Sponsor, Black Belt to work on a GB/BB project where with the inputs of everyone the DMAIC or DMADV phases to be done collaboratively To encourage participants for selecting a real life challenge which can be picked as cases to solve with the help of Lean six sigma principles

Manage by Metric: We use Key performance indicators to check the efficiency of the processes. This is something through which we can understand the process performance and think in a direction to drive continuous improvements in any industry such as Finance, manufacturing, Insurance etc. Also we can add these metrics in the excel or Power BI dashboard through which we can easily track the Profit, Rejection, Productivity etc. Manage the Metric: When we are particularly set a priority to improve a single metric to achieve a target. It gives us a short term relief but in long term there are hardly any benefit of focusing in any individual Metric. I can write a relevant example here, If we are taking more and more business from the clients, just to show our organization leadership that we have reached the target of having 'X' number of clients in this year. That's fine a short term relief to leadership. However, if we are not able to match the monthly Finished goods requirement for them. Then its a flop in CSAT numbers, no long term benefit here. How can we prevent to shift from 'Mange the Metric' to 'Manage by Metric': Training programs: Business should focus on giving proactive trainings to everyone unbiasedly which results in metrics improvement Giving priority to both qualitative and quantitative numbers: Organizations should focus on quality and quantity at the same time to improve CSAT, Product quality and long term growth. Metrics Audit in Process: There should be audit mechanism to check whether the projects selection by Continuous improvement team is done by keeping a close tap on metric current status or not. Incorporate the Customer feedback in Metrics tracking sheet: As and when we got the dump downloaded from the tool pertaining to Customer feedback, add those number in Metric sheet to track the trend of metrics. Reward the Team who contributed in improving the metrics : Recognise and give incentives to core team members so that people can critically think and watch from the lenses of improvement so that process outcome should be First Time Right . This attitude helps rigorously to improve the metrics.