Data Quality: The Issues.

Late in 2018 IHRIM produced a survey with the aim

of identifying the data quality issues which could potentially adversely affect the creation and maintenance of an electronic patient record.

The survey was initially circulated to IHRIM members but then its circulation was widened to include CIO’s and members of the BCS. (https://www.surveymonkey.co.uk/r/dataquality_4)

To say the response rate was poor is an understatement as there were only 27 responses mostly from ENGLAND (92%), with the majority >80% being IHRIM members and >90% working in the Acute Sector. Understandably, considering the responses, most (54%) were involved in Clinical Coding and Medical Records/PAS/Admin (47%) ie IHRIM members.

However, clinical data quality is an issue which affects everybody, and the wider circulation of the survey did not result in the take-up by the other professionals involved in the use or capture of clinical data. A wider uptake by the Chief Information Officers or Chief Clinical Information Officers would have produced a more significant set of conclusions.

Summary of the Survey Results

Data Quality Kitemark

Does your organisation have a data quality kitemark question resulted in a 78% ‘Don’t know’ response and only 7% giving a positive answer.

Is this an issue? Had the survey been completed by different staff, would this still have been the result? What is the data quality kitemark and how does it guarantee good quality data?

The question ‘In your organisation, how are data inaccuracies and/or duplicate records managed in the PAPER health record/casenote?’ elicited a mix bag of responses which essentially fell into the following main categories. (Number in brackets indicate where multiple responses)

  • Duplicate records are merged. If documents are in the wrong notes they are copied and put into the correct notes with a note to say they were in the wrong notes and the date they were found. The wrong document is then redacted to remove the patients incorrect demographic information and then it is left where it is to say that this has been misfiled and the date it was found and not to remove the document.
  • HEALTH RECORDS MANAGERS MARK NOTES APPROPRIATELY
  • Incident reports and through health records dept
  • Partly by Records Management Team and partly by Data Quality (5)
  • No modalities in place yet
  • These issues are managed by Medical records department and the Governance team (4)
  • Duplicate records are identified and returned to Health records Library where there is a named individual merges the records - they also then merge electronic records too. Merged and corrections made to record. Error marked with explanation and remains part of notes. They are reported and misfiled information is marked as such, duplicated so that it can be filed in the proper place and crossed out but not removed from the place they were discovered. (6)
  • Cross through with single line score through date and sign

Comment

There are various ways in which any errors in the paper record are managed and it appears this is generally a task shared between Medical Records Departments and Data Quality teams.

One Trust actually raises an incident formally so it is logged in the Trust Incident log

One worrying answer is that are no mechanisms (modalities) in place at the moment.

The question re who corrects any inaccuracies (and/or duplicates) in the digital record resulted in the following responses:

  • We merge our duplicate records on our PAS system and this then merges our electronic records.
  • The Data Quality Team manage the process as we do with the paper records (9)
  • Merge on PAS which then feeds through to the other electronic systems and merge paper notes if there are duplicates (2)
  • They are merged and correct entries added to the correct part of the record. (4)
  • Merge together, done by a combination of people • patients in our trust can have duplicate electronic record which causes major problems

Comment:

It appears most of this is done by a Data Quality Team although the detail is a little unclear in some responses. ( e.g. One response was ‘By correcting it’ doesn’t say who or how and if the incorrect data is kept in place.)

The issue is that if an error is detected, how is that error rectified and by whom. It is IMPORTANT that the incorrect data is kept in place with an obvious ‘this is incorrect’ marker attached to it, as this incorrect data may already have been used as part of clinical process.

Who corrects the data inaccuracy?

Best practice is assumed to be that the individual who makes the error must be the one to repair the damage. This can be facilitated by the Data Quality Team or by Health Records Department. >70% responses said the Data Quality Team are the ones who repair the data although it doesn’t go on to say that they fed back to the initial author to make the changes required.

Is the inaccurate data left in place or deleted?

Over 75% say the data is left in situ with a note re the inaccuracy. Note: It would be wrong to delete the inaccurate data as a clinician may have made an intervention based on that inaccurate data.

Do you currently CULL your digital data/records?

Of those who answered there were 0% who said YES. This is potentially in conflict with the current rules of retention of records.

It is interesting to note that some EPR systems are incapable of deleting data from the EPR system.

Comment: Is this unreasonable to expect staff to support the lack of automatic culling of records because the digital system will not allow it. Is this an issue? Is it illegal?

AND FINALLY…………………….what is the biggest potential barrier to implementing EPR?

  • Incorrect data
  • Staffing and money and there is not enough understanding of the new system and how it would impact different teams.
  • Users not trusting 'IT' so wanting to have paper workrounds/fallback
  • Users lack of commitment to doing data entry correctly
  • Users always believing that IT records/forms will take longer to complete and be more onerous than scribbling out something on paper (which then gets lost, misfiled, is illegible etc etc - but let's not worry about that)
  • CURRENTLY TOO MANY SEPARATE SYSTEMS, STAFF UNWILLING/UNABLE TO UTILISE ALL APPROPRIATELY
  • Clinician engagement with the project & appropriate access to electronic devices to record data.
  • Clinical involvement
  • The organisation of which records are to be done (digitised) as priority and timescale
  • Incorrect data being entered onto health record
  • Managing the changeover where the new electronic record exists alongside the paper one.
  • Fear of the unknown
  • Insufficient training and a lack of understanding about the impact of inaccurate data
  • Staff support and enough audit to ensure accuracy
  • Too many different systems to try and amalgamate and make a simpler system
  • Systems not feeding each other or a delay with data being recorded.
  • So many paper records kept in so many different departments (not all within patients main casenotes) and also the cost of implementation.
  • Technology compliance and inability of the management to sponsor it in totality
  • Information not available in timely manner, loads of different variations of pathways, training.
  • Understanding of the ground level problems is lacking when implementing the new systems
  • Organisational change is the biggest issue but also data quality when migrating data, and unearthing previous problems can be a big barrier
  • Clinical engagement is poor / Technological solution need to be more robust in making it possible • Staff resistance • Quality of scanning and patients having multiple case notes active
  • Changing culture is biggest challenge - get your clinicians on board right at the beginning and allow them to drive the implementation.
  • Change culture of staff in the NHS.
  • Cost
  • Direct entry into the system by clinical staff and having to reduce the number of patients they have time to see

Conclusions:

These potential barriers are well known and consistently stated in any survey. Clinical engagement being a common theme along with ‘fear of change/unknown, too many disparate systems, staff developing paper work-arounds because the digital solution has been poorly worked out, and finally cost.

Any business case which predicts digitising paper records will make a Trust huge savings is likely to be wrong. A full-blown, well considered and implemented digital system which is easy to use and replaces ALL the disparate systems in the Trust is NOT going to be a cheap option.

If you have not yet completed the survey, please do so at: https://www.surveymonkey.co.uk/r/dataquality_4