Twenty-six years following records to their source

Mandate

Bring clinical coding, medical records, data engineering, analytics and data governance into one accountable office.

Outcomes

• Kept health-account linkage a separate, explained step; 23 per cent first-year linkage and ₹52 lakh of incentive
• Reduced insurance deductions on comparable admissions from 6.8 to 4.9 per cent of billed value
• Assigned named clinical or operational owners to 140 data elements
• Began preparing consent notices, retention schedules and breach procedures for the Digital Personal Data Protection Act

Decisions in context

The board created this role to end an arrangement in which coding sat under finance, records under operations and analytics under technology. Bringing them together exposed how often each team had been correcting the others' figures without telling them. My first year was spent agreeing definitions and owners rather than buying tools. I report to the group chief executive and work most closely with the medical director, the chief financial officer and the general counsel. Two of my decisions in this role carried visible costs: a voluntary approach to national health accounts that reduced incentive income, and a documentation programme that slowed the revenue-cycle team's preferred solution. I present a quarterly paper to the audit committee and report twice a year to the risk management committee. I am not a director of the company, and I attend those meetings only for my own agenda items.

Reason for the move

Current appointment; considering a first independent directorship with employer consent.

Mandate

Standardise clinical coding and group reporting, and prepare patient data for a single hospital information system.

Outcomes

• Had 61,000 uncertain patient matches reviewed before the unified record went live
• Moved all hospitals to one ICD-10 coding standard with a common monthly audit sample
• Reduced the daily pandemic bed and oxygen return to six confirmed fields
• Brought the Mangaluru hospital onto group definitions at its opening in 2021

Decisions in context

This role turned a records function into a group data function. I inherited five coding teams that classified the same procedures differently, which made comparisons between hospitals unreliable and insurance queries harder to answer. We agreed one standard and audited a common sample each month. The pandemic arrived within a year. District authorities needed accurate bed and oxygen figures every evening, and wards had no time for long forms, so we reduced the return to six fields and named a confirming officer in every hospital. The largest task was preparing patient data for the new hospital information system. My insistence on reviewing uncertain matches delayed the unified record by fourteen weeks and caused real difficulty at the Mysuru opening. I would still make that call, but I would gather the evidence earlier. I reported to the chief operating officer and began presenting to board committees during this period.

Reason for the move

Appointed the group's first chief data officer when the board created a single data office.

Mandate

Bring five hospitals' medical records and management information to common definitions.

Outcomes

• Agreed group definitions for occupancy, length of stay and discharge time
• Cut retrieval time for inpatient files at the Kochi hospitals from two days to four hours
• Introduced a monthly coding audit of 300 discharge summaries
• Integrated the Kannur hospital's records at its opening in 2017

Decisions in context

I joined a company that had grown hospital by hospital, each with its own records room, registration habits and monthly report. The founding family wanted to compare hospitals before the private-equity investment, and the comparisons were unreliable. I spent my first year visiting every records department and every ward that produced a daily census. We agreed group definitions for the measures the board used most, and I published each definition beside the figure. File retrieval at the Kochi hospitals was slowing outpatient clinics, so we reorganised storage and tracking rather than requesting more space. I also began a monthly coding audit, which later shaped how we approached insurance documentation. When the Kannur hospital opened, its records followed group standards from the first day. That opening showed me how much easier it is to set a habit at the start than to correct one later.

Reason for the move

Promoted to vice-president as the company planned a single hospital information system.

Mandate

Make hospital reports consistent and usable by ward and department heads.

Outcomes

• Replaced three competing occupancy calculations with one agreed definition
• Led scanning and indexing of 4.2 lakh legacy inpatient files
• Introduced weekly ward-level reports reviewed with nursing supervisors

Decisions in context

I came to Sevanthira Memorial Hospital from the claims side, expecting to find careless records. I found careful people working with poor forms. Occupancy was calculated three ways because three departments needed it for different purposes, and nobody had asked them to agree. I spent more time on wards than at my desk, learning why entries were late and which fields nurses skipped when they were short of staff. The digitisation project was my first significant budget: 4.2 lakh paper files scanned and indexed so that doctors could see a patient's history in clinic. We indexed by admission rather than by page, after a trial showed doctors searched that way. I reported to the hospital administrator and worked closely with the nursing superintendent. The role taught me that a reporting problem is usually a workflow problem, and that ward staff help readily once a report helps them too.

Reason for the move

Joined a growing hospital group to lead records across several hospitals.

Mandate

Explain patterns in claim deductions and rejections for insurer clients and hospital networks.

Outcomes

• Built a deduction-reason classification adopted across the claims teams
• Issued quarterly documentation-gap reports to network hospitals
• Identified length-of-stay patterns that reduced manual queries on routine surgery claims

Decisions in context

A third-party administrator sits between insurers, employers and hospitals, and every one of them believes the other is the cause of a disputed claim. My work was to find out which was right more often. I started by classifying deduction reasons consistently, because the claims teams used dozens of overlapping labels. The pattern that emerged surprised me: a large share of deductions arose from missing or unclear hospital documentation rather than from policy terms. We began sending network hospitals a quarterly report of their documentation gaps. Some found it useful; others treated it as a negotiating tactic. I led six analysts from 2006 and reported to the head of claims operations. By 2008 I was more interested in why hospitals wrote what they wrote than in classifying the consequences, so I moved to a hospital where I could work on the record itself.

Reason for the move

Crossed to a hospital to work on records where they are created.

Mandate

Check incoming questionnaires and prepare survey tables for clients.

Outcomes

• Introduced back-checks on a sample of households for each investigator
• Flagged rounding patterns in age and income responses before tabulation
• Prepared tabulations for district-level household health surveys

Decisions in context

My first job was close to the source of data in the most literal sense. Questionnaires arrived from field investigators in bundles, and I checked them before anything was entered. The tidy forms taught me more than the messy ones. An investigator who rounds every uncertain age to the nearest five is not dishonest, but the result misleads anyone who later uses the table. I proposed back-checks on a small sample of households per investigator, which the research director accepted after a trial on one district survey. I had no direct reports and worked through four field supervisors, who knew far more than I did about why a household answered as it did. The surveys included health questions, and I became curious about how health data was used once it left a survey. That curiosity took me to insurance claims in Bengaluru.

Reason for the move

Moved to Bengaluru to work with health-insurance claims data.