Blind Spots in Hospital Data

An interactive analysis of challenges in collecting data for Culturally and Linguistically Diverse (CALD) communities in Victorian cancer services.

The Problem: A Significant Data Gap

The core issue identified is a stark difference in the completeness of patient records. Patients from non-English speaking backgrounds are significantly more likely to have incomplete data, creating "blind spots" in our understanding of their health needs.

View Incomplete Data Rates By:

Country of Birth Preferred Language

This chart visualizes the percentage of patients with one or more 'unknown/indeterminate' answers in their CALD data fields.

The Investigation: Why Does This Gap Exist?

A mixed-method approach was used to understand the root causes of these data gaps. Explore the key findings from different facets of the investigation by selecting a category below.

Staff Perspective Systems Review Community Voice

Lack of Training and Resources

A key finding from staff surveys and interviews was that administrative staff, who are primarily responsible for patient data collection, receive no formal training in CALD data collection. This leads to a lack of confidence and consistency when asking sensitive questions about a patient's cultural and linguistic background.

System and Process Limitations

The review of IT systems and patient journey mapping revealed limitations. For instance, lists of languages often do not include relevant dialects, making it difficult to capture true linguistic diversity. Furthermore, robust follow-up mechanisms to correct or complete missing data are not consistently integrated into standard workflows or IT systems.

Mismatch with Community Identity

Focus groups with CALD community members highlighted that standard data variables do not always capture important aspects of identity. Variables such as 'year of arrival in Australia', 'ethnicity', and 'culture', while not part of the minimum standard set, were identified by communities as crucial for a more complete representation of diversity.

The Solution: Recommendations for Clearer Vision

Based on the investigation, several key recommendations were proposed to address the data gaps and improve the quality of CALD data collection across the health service.

Implement Staff Training

Provide training on sensitive data collection strategies to build staff confidence and improve the quality of information collected with cultural and linguistic sensitivity.

Develop CALD-Specific Resources

Create and distribute resources for data collection specific to CALD populations, similar to those available for Indigenous Status questioning.

Establish Robust Follow-Up

Integrate follow-up mechanisms in both standard work procedures and IT systems to enhance the completeness and accuracy of data.

Incorporate Clear Guidelines

Update internal systems and policies (e.g., within CanMAP) to specify mandatory CALD data collection requirements clearly.

Enhance Language Options

Ensure that lists of languages available in data entry systems include relevant dialects to better capture linguistic diversity.

Expand Standard Variables

Consider including 'ethnicity', 'culture', and 'year of arrival in Australia' to better capture diversity beyond language and country of birth.

Interactive report based on the poster: "Blind spots in the hospital: an absence of CALD data" by M. Iserlohe, M. T. Mok, V. Joshi, U. Hafeez.

document.addEventListener('DOMContentLoaded', () => { const cobData = { labels: ['Country of Birth: MESC*', 'Country of Birth: non-MESC*'], datasets: [{ label: '% of Patients with Incomplete Data', data: [13, 24], backgroundColor: ['rgba(148, 163, 184, 0.6)', 'rgba(15, 118, 110, 0.6)'], borderColor: ['rgba(148, 163, 184, 1)', 'rgba(15, 118, 110, 1)'], borderWidth: 1 }] }; const langData = { labels: ['Preferred Language: English', 'Preferred Language: non-English'], datasets: [{ label: '% of Patients with Incomplete Data', data: [15, 32], backgroundColor: ['rgba(148, 163, 184, 0.6)', 'rgba(15, 118, 110, 0.6)'], borderColor: ['rgba(148, 163, 184, 1)', 'rgba(15, 118, 110, 1)'], borderWidth: 1 }] }; const chartConfig = { type: 'bar', data: cobData, options: { responsive: true, maintainAspectRatio: false, indexAxis: 'y', scales: { x: { beginAtZero: true, title: { display: true, text: 'Percentage (%) with Incomplete Data', font: { size: 14 } } }, y: { ticks: { autoSkip: false, maxRotation: 0, minRotation: 0, callback: function(value, index, values) { const label = this.getLabelForValue(value); if (label.length > 20) { return label.split(' ').map(word => word.length > 10 ? word.substring(0, 10) + '...' : word); } return label; } } } }, plugins: { legend: { display: false }, tooltip: { callbacks: { label: function(context) { return ` ${context.dataset.label}: ${context.raw}%`; } } } } } }; const ctx = document.getElementById('completenessChart').getContext('2d'); let completenessChart = new Chart(ctx, chartConfig); const chartTakeaway = document.getElementById('chart-takeaway'); const takeawayCobText = `Patients born in non-Main English Speaking Countries (non-MESC) are almost twice as likely to have incomplete records (24% vs. 13%). <br> <em class='text-xs'>*Main English-Speaking Countries</em>`; const takeawayLangText = `The disparity is even greater by language: patients whose preferred language is not English are more than twice as likely to have incomplete records (32% vs. 15%).`; chartTakeaway.innerHTML = takeawayCobText; const viewByCobBtn = document.getElementById('viewByCob'); const viewByLangBtn = document.getElementById('viewByLang'); const navButtons = [viewByCobBtn, viewByLangBtn]; function updateChart(data, takeawayText, activeButton) { completenessChart.data = data; completenessChart.update(); chartTakeaway.innerHTML = takeawayText; navButtons.forEach(btn => btn.classList.remove('active')); activeButton.classList.add('active'); } viewByCobBtn.addEventListener('click', () => { updateChart(cobData, takeawayCobText, viewByCobBtn); }); viewByLangBtn.addEventListener('click', () => { updateChart(langData, takeawayLangText, viewByLangBtn); }); const tabButtons = document.querySelectorAll('.tab-button'); const tabContents = document.querySelectorAll('.tab-content'); tabButtons.forEach(button => { button.addEventListener('click', () => { tabButtons.forEach(btn => btn.classList.remove('active')); button.classList.add('active'); const tabId = button.dataset.tab; tabContents.forEach(content => { if (content.id === `tab-${tabId}`) { content.classList.add('active'); } else { content.classList.remove('active'); } }); }); }); });