On Monday, the NHL released a memorandum describing its protocol for phase 2 of its return-to-play plan. The 21-page document lays out how the NHL will bring players back to their teams’ cities and protect their health as they begin training again.

It is an unserious plan that will put lives at risk.

First, a disclosure: I am not a medical, public-health, or infectious-disease expert. I don’t talk much about my day job here, but I have more than fifteen years of experience in clinical research as well as global frameworks and regulations for quality, privacy, and safety. Also, in reviewing the NHL’s plan and writing this story, I consulted with Aileen O’Hearn, my partner, who has a PhD in microbiology and immunology and more than ten years of direct experience with infectious diseases.

Here is the full text of the memo.

Travel and Housing

In Phase 2, the league ends self-quarantine and expects players to return to their club cities to begin conditioning. As a policy-level statement this is reasonable, but it becomes immediately problematic:

Clubs should help to shall facilitate Player travel arrangements, to the extent permitted, to enable Players who are not in the Club’s home city, to return to the Club’s home city as each Player may deem appropriate, in order to facilitate their ability to engage in Phase 2 activities.

This is the protocol’s first delegation of responsibility to clubs, and it’s riddled with qualifiers — and a typo. The language is weasel-y; “should”, “help”, “shall”, and “facilitate” all muddy what exactly the teams are supposed to do here. And get ready to see “to the extent permitted” or “to the extent possible” a lot in this plan.

On the other hand, the agency granted to the player (“deem appropriate”) is good. That implies that players get veto power over whether they want to return at a certain time. If I were a player in rural western Canada, I would not want to fly to the DC Metro area right now.

To that point, the implicit expectation is for players to take charter flights, though their $1,500 reimbursement would not nearly cover that . Public transportation would require a 14-day quarantine, and carpooling is explicitly forbidden. The burden for coordinating and paying for these flights is vague. I’d presume whatever the equivalent of carpooling is for charter flights would also be forbidden, though this is not clear. So we will likely have, to save money, five or six Swedes climbing into the same private jet next week.

Here’s a line I worry about:

To the extent travel and shelter in place restrictions impede or delay the ability of your Players to return to your Club’s home city, please notify one of us so that the League stays informed on the challenges each Club may be facing in this regard.

This line sets a pattern that is repeated throughout the protocol: an expectation of smooth success without an escalation plan for failure, and an non-specific responsibility to notify the league without the establishment of a central reporting structure for safety or updates to the protocol.

Once in their cities, the players without homes or apartments will be put up in hotels. Good ones too:

The accommodations must be of the same high quality provided to Players during the Regular Season. The hotel or other accommodations shall satisfy the hygiene, distancing, cleaning and disinfecting requirements recommended by the Club’s medical professionals, including its infectious disease consultant.

There is no framework by which these hotels will be qualified for their hygiene, distancing, cleaning and disinfecting requirements beyond the approval of the Club’s medical professionals, which includes a position I am not familiar with: the “infectious disease consultant.”

We do not know the qualifications of these persons. Are they supposed to experts? If so, ID experts are presumably in short supply right now; I’m not sure how they could be hired. Are the clubs responsible for qualifying them and hiring them? Even the Ottawa Senators? Once hired, I’m not sure how they could be empowered to unilaterally shut down a club’s progress towards restart without undue pressure being put on them.

Testing

This paragraph does a lot of work:

As an over-riding principle, testing of asymptomatic Players and Club personnel must be done in the context of excess testing capacity, so as to not deprive health care workers, vulnerable populations and symptomatic
individuals from necessary diagnostic tests (“Publicly Necessary Testing”).

Players are typically healthy and are therefore less likely to be tested, especially in a country where regular universal testing is not yet possible. The NHL justifiably does not want testing of players to deprive others who need it more, but — less justifiably — creates a convenient mechanism to circumvent the testing protocol they’re about to describe. Test shortage is baked in, so there’s an excuse not to test your players.

The alternative would have been to define a measurable minimum testing threshold necessary before club activities can begin again, and have teams demonstrate that they can meet that threshold before they are given approval to start up. This is another example of non-specificity and delegation to the clubs without any real oversight or central reporting.

The tests themselves come in three layers:

• A return-to-play PCR test
• Twice-weekly PCR tests, if possible
• Daily self-administered temperature and symptom tests

The return-to-play test can be skipped based on the “over-riding principle” stated above. The league has a good contingency here though: a mandatory 14-day self-quarantine period if the test cannot be done.

The twice-weekly tests are similarly qualified and given an extra carve-out (“if possible”), further tempting club non-adherence.

The tests themselves are left stupefyingly vague. Clubs are apparently responsible for procuring and qualifying the tests — with the understanding that tests are not readily available, especially for asymptomatic and otherwise typically healthy non-elders. There are plenty of tests on the market that do not have FDA approval. Are these acceptable? There are more tests that have been given FDA fast-track approval, but these tests come with the risk of lowered accuracy and sensitivity. Can those tests be used?

The league offers another exception here: serology tests, which may alternately detect presence of prior or current antibodies (and therefore prior or current infection), can be used. The league helpfully says that FDA emergency-use authorization is not sufficient for these tests, but does not state which kind of detection is appropriate.

Self-testing and self-reporting by players is, on its face, absurd.

Independent and disinterested oversight is critical, and even that oversight fails with the understanding that many disease-spreaders are asymptomatic. The league has concussion spotters for a reason, and that protocol is still notoriously flawed.

Self-testing is unenforceable. The testing rubric is insufficient for the disease’s profile. The players are supposed to report to club professionals whose qualifications are unclear and whose availability is uncertain. The self-testing is augmented by twice-weekly testing, except when it isn’t, and the reliability of those tests has not been verified. Got it.

And then this: The protocol has no procedure for what to do when a player tests positive.

This is the entirety of it:

During Phase 2, anyone who develops symptoms (or if persons sharing a home develop symptoms or tests
positive for COVID-19) shall immediately notify Club medical staff of such, shall self-isolate, and shall be medically evaluated by the Club’s physician(s), who shall consult with the Club’s infectious disease specialist to determine next steps, and administer PCR testing, if appropriate.

This is a non-standardized and ad hoc plan for escalation, again referring to the club’s infectious disease specialist, who may or may not exist. And the plan is entirely restricted to the infected player. The league delegates all contact-tracing responsibilities to the club, but offers no description of resourcing or training on the matter. The league does not require that club activities completely or even partially stop. The league does not describe responsibilities for the housing of and quarantining for infected persons.

The league does require notification of infection, though the language of the memo implies this is for CBA-related purposes:

[. . . ] the Club Physician shall issue an Exhibit 25-A designating the Player as “unfit to play”, the Player shall be deemed to have sustained an illness arising out of the course of his employment as a hockey player for such period as he may be removed from training, practice or play, and his condition shall be treated as a hockey related injury for all purposes under the Collective Bargaining Agreement, unless it is established, based on the facts at issue, that the Player contracted COVID19 or the resulting or related illness outside the course of his employment as a hockey player.

From the text of the memo, it seems that public disclosure of infections is not required.

Hygiene

The plan for practicing is straightforward and good, relatively speaking. Workouts are limited in the number of their on-ice and support personnel. The workouts themselves are no-contact, primarily intended for conditioning so players are safe to compete in the future. Players are required to wear masks when not working out. Support personnel wear PPE and are split into those with and without close access to players. Deep cleanings of equipment and facilities between uses by groups is required.

But there are a lot of gaps. The league provides a carveout to reuse support staff between workout groups, which defeats the entire point of insulation (emphasis mine):

During Phase 2, Clubs must, to the extent possible, assign a unique Athletic Trainer, Strength and Conditioning Coach, and Equipment Manager, respectively, to each group of six (6) Players, so as to limit cross-exposure among groups. To accomplish this, it may be desirable to involve additional personnel, such as those from the Club’s AHL affiliate.

The league is effectively asking for teams, one of which is the Ottawa Senators, to spend extra money on their staff to avoid cross-contaminating players.

Further, clubs are presumed to have a large stock of industrial cleaning supplies on hand, as well as staff to perform these cleanings. If they do not, they are asked to voluntarily notify the league:

Clubs who have concerns about their ability to obtain sufficient amounts of cleaning and disinfecting solutions shall contact the League promptly.

This is another delegation of responsibility without any proactive oversight. It is an invitation for noncompliance.

Regarding the no-contact nature of the workouts, I hope this is not considered a preventative measure of any significance. Athletes who are working out and breathing heavily and in close quarters — be it lined up next to each other for skating drills or taking faceoffs a few inches apart — might not be any safer from infecting one another than they would be making full contact with one another. Effectively what the league is doing is implementing isolation pods, where the players and their close contacts are almost certain to expose one another, but they as a group are insulated from other pods. But that works only as long as the clubs conduct regular extensive cleaning and reuse no support staff across groups.

Conclusions

This is not a real plan. This is a blue-sky-sunny-day description of what it would be like if nothing were to go wrong when emerging from a pandemic. In 21 pages, there is one sentence about how to care for a player who tests positive.

I am surprised that the NHL put this together in coordination with experts.

The major problems:

• Delegation to clubs. The league has dumped onto the teams endless responsibilities: hiring and qualifying ID experts, procuring and qualifying tests, staffing up support personnel and cleaning workers, obtaining large amounts of cleaning supplies, conducting contact tracing, and literally all healthcare escalation. The league does not compel the teams to certify that they have done or are doing any of these things. There is no description of independent review anywhere in here.
• Too many carveouts and exceptions. I counted at least ten supposed requirements that baked in open-ended excuses like “to the extent possible.” The league acknowledges that a one-size-fits-all solution won’t work for everyone, but there are no controls around these exceptions. In a situation in which certain resources are already limited, and in which the material and financial situations of teams are variable, this plan is designed to fail. There are no evaluations or measurements that can be used objectively, nor any mechanism by those measurements can be compared to each other to improve the protocol across all 31 teams.
• Ambiguity of qualified staff. This plan needs lots of competent and hardworking people to keep players safe, but there are not controls describing their competencies. It’s not clear how infectious disease experts are obtained by clubs or how they are deemed experts in the field. It’s unlikely that there are many of these persons with free time right now, and there are no assurances that their judgment is independent and free of outside influence. Given the NHL’s historical troubles with health professionals, it would be naive to think this situation would be any different.
• No oversight. At all. There is no requirement for teams to document how they’ll meet this protocol, no central review process for those plans, no reporting structure beyond CBA requirements, no plan to revise and improve the protocol  based on feedback. Teams are asked to tell the league when they cannot do something as prescribed, but there are no milestones or measurements to anchor anything. This plan is based entirely on trust.

This protocol is supposed to protect players and their families, but it is riddled with holes. Its obvious risks are not mitigated. There is no accountability for its execution. There is no way to trace that it is being followed.

I hope it works.