Data analytics may help prevent patient no-shows for imaging

2015 12 08 12 51 57 742 Patient Hispanic 200

Missed patient appointments can have a significant impact on patient care and cost radiology departments millions of dollars in lost revenue. Data analytics may be able to help address this problem, however.

Researchers from Massachusetts General Hospital (MGH) have applied analytics to identify the patients who are most at risk of missing their radiology appointment based on various sociodemographic and economic factors. Their findings pave the way for interventions to decrease no-shows.

"We need to leverage technology to engage our patients and ensure that no patient is left behind," said presenter Dr. Omid Khalilzadeh, now a radiology resident at Mount Sinai Beth Israel in New York City.

He shared the research team's experience with analytics during a scientific session at the RSNA 2015 meeting in Chicago.

No-shows

As is the case with other specialties, radiology departments experience a considerable number of patient no-shows for appointments. These no-shows have both a financial impact and a negative effect on patient care, Khalilzadeh said. It's been uncertain, however, what factors contribute to missed appointments.

Since the term "no-show" places the responsibility for the missed appointment solely on the patient, the researchers decided instead to use "missed care opportunity" to also account for the provider's responsibility in ensuring that the patient has access to care, he said. A missed care opportunity was defined as a healthcare encounter that was scheduled but not attended.

"Each missed radiology appointment is a missed care opportunity," he said. "Each missed care opportunity results in delayed diagnosis and negatively impacts patient care."

The researchers hypothesized that missed care opportunities in radiology may be a symptom of missed care in other specialties and, more generally, disparities in medical and radiology care, Khalilzadeh said.

As a result, the researchers set out to delineate the risk factors associated with missed care opportunities and quantify a patient's access to medical care via a model they called a patient connectivity index. Ultimately, they hoped to use this index, which accounts for various sociodemographic and economic factors, as a model for identifying those most at risk of a missed care opportunity.

The researchers retrospectively reviewed more than 1 million imaging studies between February 2014 and February 2015. Of these, there were 551,000 scheduled outpatient studies, with 42,000 (8%) resulting in missed care opportunities. These missed appointments were attributed to 28,000 unique patients.

Leveraging both internal data sources such as the institution's RIS and longitudinal medical record with external resources such as Google Maps and Boston city census data, the team utilized machine-learning techniques and multivariate, multilevel regression analysis to create the patient connectivity index, which takes into account factors such as a patient's age, gender, distance from hospital and transportation, insurance, primary language, ethnicity or race, the time of year, billing codes, and referral pattern.

Risk factors

Of the patients who had a missed care opportunity, 18,000 were females and 10,000 were males; 24,000 were English-language speakers, while 5,000 were non-English speakers. Finally, 21,000 were of white ethnicity and 7,000 were of other ethnicities.

The data analysis revealed several risk factors that had a statistically significant association (p < 0.01) that a patient appointment would result in a missed care opportunity:

  • A lower educational level
  • Lower income
  • Primary language that wasn't English
  • Living out of state
  • African-American (odds ratio of 1.8 compared with white) and Hispanic (odds ratio of 1.5 compared with white) racial backgrounds

The researchers bundled the risk factors into the patient connectivity index, which on a scale of 1 to 100 quantified the patient's access to healthcare. A connectivity score of 1% to 20%, for example, indicated that a patient had a low access to care and, thus, was more likely to miss an appointment.

"If the patient is high risk for having low access to care, the aim is to implement actionable items to improve their access," Khalilzadeh said.

The research shows that socioeconomic disparities exist in radiology, he noted.

"Missed care opportunities result in direct and negative impact on patient care and mostly affect minorities -- like non-English primary language speakers, low-income households, and non-White race -- who are more vulnerable to poor medical care," Khalilzadeh said.

Missed care opportunities per radiology specialty
  Number of missed care opportunities Revenue loss associated with lost Medicare payments
Musculoskeletal 15,493 $1,885,482
Gastrointestinal/genitourinary 7,791 $2,223,459
Breast imaging 7,699 $658,557
Thoracic 5,284 $983,666
Cardiovascular 2,401 $617,955
Neuroradiology 2,132 $940,910
Nuclear medicine 658 $680,018
Interventional radiology 314 $489,770

Bridging the gap

With this information in hand, radiology departments could, for example, implement a platform for assessing the risk of missed care opportunities for each patient scheduling an appointment. They could also bridge the language gap by providing reminders in the patient's primary language and also address economic concerns and provide transportation aid, Khalilzadeh said. In addition, they could also offer easy ways to reschedule appointments.

"The goal is to diversity healthcare delivery and focus on equitable care to improve access," he noted.

These efforts could also yield a substantial financial return. As part of its research, the MGH team found that missed care opportunities had a significant financial impact on its radiology department over the study period.

Page 1 of 603
Next Page