Cloud Hosting

Youre reading the web edition of STAT Health Tech, our guide to how tech is transforming the life sciences.Sign up to get this newsletterdelivered in your inbox every Tuesday and Thursday.

Elizabeth Holmes aims to undercut prosecutions claim of deception

Elizabeth Holmestestified in theTheranosfraud trial Monday that the companys blood-testing technology performed well in early studies with drug makers such asMerck,Astra Zeneca, andBristol Myers Squibb. The line of questioning by the defense fielded with slow, steady replies from Holmes was meant to rebut testimony by prosecution witnesses suggesting the company sent falsified reports on validation testing by drug companies to investors. Holmes also countered testimony from aPfizerdirector, Shane Weber, who told jurors that Theranos replies to his technical questions were at times deflective or evasive.

advertisement

By calling Holmes to testify, her lawyers are betting that she will be able cast doubt on prosecutors narrative that she deliberately misled investors and the companys potential clients. The big question is how she will hold up under cross examination.

Biotechs big data problem is all too human

advertisement

As biological datasets have exploded in size, biotechs need data scientists to process their plentiful petabytes. But many are struggling to hire enough researchers with the computational know-how needed for the job. Executives believe the gap might [become] fourfold unless they start acting now, Parag Patel, a partner atMcKinsey and Companyfocusing on the life sciences industry, toldSTAT. Part of the problem is supply and demand: There are only so many top-tier data scientists, and pharma is competing for them against companies likeGoogle,Apple, andFacebookthat cannot only pay more, but are often more data-proficient, too. It can get frustrating just cleaning up dirty data, said Neal Cheng, a data scientist who left biotech to work at eBay.Read more about the biotech brain crunch in our colleagueAngus Chens new story.

Verilys clinical trials business leans into data

At theSTATSummitlast week, Katie spoke withVerilys leaders about their plans for their clinical trials infrastructure, just one arm of the life science companys sprawling and sometimes unfocused business.Amy Abernethy, who joined theAlphabetsubsidiary over the summer after championing real-world data and evidence at theFDAas principal deputy commissioner, described her next steps as president of clinical trials platforms. Practically speaking, we are going to focus on building longitudinal datasets that can be put to use for understanding the contours of illness, said Abernethy, adding that such data could be used to test multiple different treatments simultaneously. Abernethy said the company hopes to create a vast repository combining real-world data, medical records, and traditional clinical trial data. Mario has moredetails.

The dream of data-sharing in 2030

By the end of this decade, publichealthresponse and preparedness will be driven by access to real-time data. In this day and age, that seems like a fairly basic desire. But it is among many distant goals that made it onto anew interoperability wish listcompiled by theOffice of the National CoordinatorforHealthIT. To help guide its policy roadmap, ONC asked providers and other stakeholders to list outcomes they hope to achieve through enhanced data interoperability by 2030. The agency received more than 700 responses. Among the other priorities on the list: Patients and doctors will be able to compare the costs of drugs, tests, and procedures online before the bill comes in.

Epics sepsis alerts skyrocketed during the pandemic

A general rule for AI models is that they should be trained on data from the kind of patients they are likely to see in real clinical settings. But what happens when the AI encounters a fundamental change, like the onset of Covid-19? Anew studyof a sepsis algorithmdeveloped byEpic Systemsfound that the number of alerts it generated in two dozen hospitals jumped 43% in the first three weeks after Covid-19 hit. The study by researchers at theUniversity of Michigandid not measure the accuracy of the alerts, but pointed out the increased volume may contribute to alert fatigue and prove detrimental at a time of constrained resources. The finding also points to the need for careful monitoring and governance of AI algorithms after they are implemented.

Meanwhile, clinical predictive tools continue to be adopted around the country, asnew survey datafrom theCollege of Healthcare Information Management Executivesshows. In the last year, the percentage of surveyed acute care organizations with predictive tools integrated into clinician workflows increased from 44% to 55% using a mix of homegrown tools along with those from EHRs and third parties. Among ambulatory care providers, adoption was at 52%, and at 40% in long-term post-acute care facilities.

Athenahealths acquisition & moves for AI drug discovery

Movers & shakers

What were reading

See the original post:

Elizabeth Holmes testifies in Theranos trial & Athenahealth acquisition 2.0 - STAT

This is a guest blogpost by Peter Jackson, Chief Data and Analytics Officer at Exasol.

The growing mainstream appreciation for quality data and its resulting insights has made the Chief Data Officer (CDO) role more acclaimed than ever. There is now a CDO in almost two thirds (65%) of companies a huge jump since 2012 (12%).

The CDO understands the processes and drivers behind success based on data making them distinct from the CIO. They provide a central point of responsibility and accountability for business clarity, efficiency, and performance and play an increasingly important role in shaping and influencing the strategic direction of a company.

As this role evolves, Exasol spoke to 250 active CDOs from across the UK, the US and Germany to find out more about these individuals and their experiences.One promising finding is that the role is opening C-Suite doors to women who have been overlooked for senior executive roles for many decades. Of the CDOs we spoke to, 26% identified as female and 73% male slightly bucking the IT industrys gender talent gap. For example, when looking at the wider gender gap found across the technology workforce, 19% are female, whilst for data science roles in general, as few as 15% are women.

Commenting on this trend, talent specialists for C-Suite appointments, Savannah Group, told us at Exasol, Organisations love to hire female talent. There are some incredible female CDOs and they are able to command a premium and cherry pick and shape the role more. Its an amazing time to be a female CDO!

Another surprise we uncovered in our research is that todays generation of CDOs are strikingly young. According to M&A Executive Search, the average age of a typical C-suite member is 56, yet the majority of the CDOs we spoke to (55%) were between the age of 25-34 when they were first appointed to the role. Interestingly, 18% of the German respondents we spoke to were between 18-24 when they took on the role more than double the number from the UK and the USA that were in that age bracket.

This more diverse pool of talent is encouraging as its not always easy for companies to find candidates. UK government research found that almost half of businesses (46%) have struggled to hire for roles that require data skills.

The opportunity for change

Diversity isnt just about demographics, however. Its also about having a diverse skillset, which is increasingly important in tech, particularly in data science.

As you might expect, 73% of the CDOs we spoke to had come from a technical background, and 69% say that most people in their teams are from a technical background too. But businesses suffering from tunnel vision and only hiring CDOs from purely technical backgrounds are potentially missing out.

Diversity is crucial when liberating data, as well as improving data accessibility and literacy across all levels of the organisation. If we concentrate too muchon STEM in education, then well have people who understand the numbers and stats, but they wont necessarily have the skills to articulate the meaning of that data, turn it into actionable insights or communicate next steps around those insights to the business.

Derek Danois, CDO at GE Healthcare, for example, has seen real benefits from employing a diverse team. I have a combination of people on my team. A physician, someone with a healthcare IT background and another with marketing background and MBA, he said. This makes for great conversation between them and a cross pollination of ideas. A marketing background is a good thing as we have to communicate across a number of business functions, and they are a strong liaison to the commercial part of the business to gather the data analysis they need. And they are someone that is credible at the table, for example, when discussing customer engagement.

The CDO role is no longer simply about governing data its about liberating it. Transforming data from a confusing, gated asset into an open, useful tool, accessible and actionable to everyone is realised on the top line (revenue) and bottom line (cost reduction and efficiency).

Its exactly for this reason that it will be exciting to see the impact that this generation of CDOs, and the next, will have on business as we know it as organisations open their minds to new candidates.

Go here to see the original:

Who are the current generation of Chief Data Officers? - ComputerWeekly.com