DATA QUALITY

Introduction

The importance of data quality

The success of the Case Management System (CMS) and Coronial Information System is highly dependent upon the Coronial Services' ability to deliver high quality data that is timely, comprehensive, reliable and complete. The aim of the National Office is to identify best practice as well as deficiencies in the quality of the data held within CMS.

CMS’s quality is both a local and national responsibility, with regional offices and the National Office taking steps and procedures to ensure that any deficiencies are identified and rectified within a timely manner and also to ensure that improvement processes are implemented and maintained.

The main purpose of the CMS data quality activity is to discover any inconsistencies and other anomalies in the data and recommend data cleansing to improve the quality of the data maintained in the information system.

Quality Objectives

Maintain systematic quality analysis and monitoring through the role of the System Administrator
Provide coder and data input training to all CMS users
Review and update CMS and the Coronial Services Unit’s procedure documents
Provide Help-desk assistance through the System Administrator and CMS super users
Review coding classification and ensure the codesets reflect current international standards and concepts
Maintain contacts within the Coronial Services Unit to ensure consistency throughout the country
Maintain contacts with external agencies also providing core injury related data
Review CSU coding classification procedure and ensure the codesets and coding applications reflect current international standards and concepts.

Principles

Dimensions of Data Quality

1. Completeness

Verification that all mandatory data fields are complete

CMS functionality has ensured that a case can not be closed unless all mandatory data fields have been coded.
All applicable data fields coded (includes mandatory and optional fields).
Coding is done using the appropriate codeset and in the appropriate format.

2. Comprehensiveness

Verification that the coding gives a complete representation of who, when, where, how and why the deceased was injured and died

Using available documents, such as the post mortem report, police investigation summary (pol47), witness statements, medical records and coronial finding, the National Office Quality Team ensures that the necessary information is coded appropriately, this includes details of the:

basic demographic data of the deceased (ethnicity, socio-economic factors etc);
date, time and location of the incident and death;
circumstances of the death – summarised in the location event description;
activity the deceased was involved in at the time of the incident;
how the deceased was injured, mechanisms and objects or substances involved;
other contributing factors and risk factors, mechanisms, objects or substances.

3. Relevance

Verification that only contributing or indirect but somewhat related mechanisms, objects, modes of transport etc are coded

Using available documents, such as the autopsy report and coronial finding, the National Office Quality Team ensures that only contributing mechanisms, objects, substances are included in the coding of the case.

Any injuries sustained after death or that do not contribute to death are not to be coded in CMS.

For example: post mortem determines that the deceased died from the blunt force injuries resulting from the initial impact of a motor vehicle crash and was deceased before the vehicle burst into flames, no evidence of breathing smoke, ash or fire. There is no need to code the contact with fire, flame and smoke as this contact did not contribute to the death and is therefore irrelevant to the cause of death.

4. Reliability

Verification of consistency in national coding, ensuring all regional offices are comparable to national requirements

The use of programmed edits and protocols ensures that the same assessment is made of every case audited by the National Office Quality Team, regardless of the region, case type, date of assessment or auditor.

The recommendations and amendments requested by the National Office Quality Team are generated from a pro forma that ensures that each regional office is alerted to an error in a similar manner, with deviations being made only to allow for case differences.

The Coding must be reliable and consistent across and within all Coronial Services’ regional offices. Often the results of a search are dependent on the consistent coding of a specific mechanism or object. For this purpose we must ensure that all cases involving similar mechanisms of death are coded the same way.

For example: All gunshot wounds are coded as:

Mechanism of Injury – Level 1: Piercing, Penetrating Force

Mechanism of Injury – Level 2: Puncturing, Stabbing

Mechanism of Injury – Level 3: Shot by firearm

Consistent coding ensures that one search of the mechanism of injury data fields will return all deaths involving a gunshot wound.

5. Validity

Verification that coding is a true reflection of the case file, reports and coroners findings

Each electronic coronial case is audited at National Office on closure based on the information available at the time of the audit. Available documents are limited to those attached to a case file which will include the:

pol47;
post mortem report;
medical history;
witness statements;
toxicology report; and
Coroner’s Finding.

Data Entry

Data entry of the information into CMS is performed locally in each regional office. In most cases the data is entered directly into CMS around the time that the death is reported. CMS was originally designed as a court information system for the Ministry of Justice.

Consistency of Source Documents

The data entered into CMS is collected from sources such as the police investigations, post mortem reports, toxicology reports and Coroner’s Findings. It is acknowledged that the primary intention of these documents is for the death investigation and legal concerns regarding the circumstances and causes of death. As a result there may be limitations when these documents are considered for statistical use.

Coding Reference

Data Elements

All CMS data elements will be subject to regular scrutiny to ensure that where the data item is based upon an external source (NCIS Data Dictionary or International Classification of External Causes of injury) that the most current version of this source data is being used.

Regular reviews of the data domains in use will also ensure that the data being produced is robust, reliable and comparable to similar national and international statistical collections.

Coding Schemes

The use of a national and international compatible coding scheme allows for the data held within CMS be compared with death incident data held in countries throughout the world.

The majority of data items are based on the definitions outlined in the NCIS data dictionary and additional fields requested by core external stakeholders. This ensures that fields such as Date of Birth, Sex, Ethnicity, Country of Birth, Sexual Orientation and Marital Status are consistent with other data collections.

A classification based on the International Classification of External Causes of Injury (ICECI) is used for the coding of:

Activity;
Context;
Counterpart;
Intent;
Location;
Mechanism of Injury;
Mode of Transport;
Objects or Substances Producing Injury
Sports and Recreation; and
User.

In certain instances it was necessary to modify the ICECI to ensure that some of the codes including information specific to Australia, New Zealand and coronial cases.

Coding Guidelines

There are three main documents available; each provides information on either the core CMS data fields or an explanation of the coding principals.

1. NCIS Data Dictionary

The most recent version of the NCIS Data Dictionary is Version 2 (July 2007). New Zealand is submitting datasets specific to the New Zealand context for the Data Dictionary update in 2010. The Data Dictionary provides a comprehensive listing of most of the CMS data fields and includes a:

uniform definition and explanation of each data field;
statement of the type of information required; and
full description of the classification (codeset) to be used for applicable data fields, including user guidelines.

NCIS Data Dictionary [ PDF 1123KB ]

2. NCIS Coding Manual

The Coding Manual provides a step-by-step explanation for the coding and completion of a selection of data fields. The manual provides practical instructions, examples of specific scenarios and guidelines for the entry of more complex and difficult core data items.

Although designed as a guide only, the manual is intended to be used as a daily reference tool by those responsible for the entry and coding of coronial cases in NCIS and CMS.

It should be noted that due to the case variations of coronial cases, the manual cannot provide instruction on how to code each individual coronial case but does offer examples and suggestions on how a particular or similar scenario would be coded.

NCIS Coding Manual [ PDF 5773KB ]

3. Case Management System (CMS) User Manual

A more technical manual which provides step-by-step information on areas relevant to CMS use, areas discussed include:

logging in and using passwords;
adding a case to the system;
signing off a case;
entering Coroner’s orders
entering requests for information
finding and editing a cases; and
administration features.

CMS Coder Training

CMS Coder Training has been conducted in all jurisdictions. During training relevant and available staff attend sessions that provided trainees with detailed information and explanation relating to many of the CMS data fields as well as allowing time for assisted data entry of real coronial cases. Additional training sessions will be conducted as required.

Data mortality coding

All Coronial Cases post 1 July 2007 are coded to the World Health Organisation’s International Classification of Diseases (ICD 10) and International Classification of External Causes of Injury (ICECI).

Additional CMS functionality flags Supplementary Classifications

Quality Review Activities

Quality review activities are not restricted to the National Office Quality Team audits on closure. The Systems Administrator conducts various quality activities throughout the data entry and coronial inquiry process. However, open case quality activities are hindered by resource and time constraints.

Local Quality Processes

The majority of data fields collected in CMS are controlled by a dataset which is specific to each data field and limits the choices or options for the field (for example, Sex is limited to Male, Female, Unlikely to be Known). Drop-down menus are used to make data entry as easy and accurate as possible. Where applicable free text data fields are used, these fields allow coders to enter case specific information which cannot be explained using the standard codeset.

Automated edits and warnings are incorporated into CMS to ensure that mandatory data fields are completed prior to the closure of the case. These edits help to ensure that the minimum of information and coding is completed before the case is closed off.

A more thorough review of each case is conducted by the National Office Quality Team upon closure of the case. Further quality checks are performed by the NCIS Quality Team, prior to New Zealand’s coronial information being uploaded onto NCIS.

Quality Assurance Audits

All coronial and advice cases are subject to monthly audits performed by the National Office Quality Team by running reports that identify errors
All closed coronial cases are reviewed by the National Office Quality Team
From mid 2010, all closed coronial cases will be reviewed by the NCIS Quality Assurance Team.

Document Actions

Print this

Coroners Court