Mood Tracking: My System for Reducing Psychiatric Hospitalizations

During my first psychiatric hospitalization in 1998, I was strapped down, placed in 4-point restraints, and administered a painful catheter—apparently because I had peed on the floor during the course of my psychotic episode. A nurse was assigned to watch over me. I was terribly uncomfortable, and over and over I begged her to release me. Finally, after a considerable time, I was released.

This blog is not about the cruelty of psychiatric institutions, although they can be cruel. Being kept in a locked ward is, in and of itself, a wretched experience, and we should work toward its elimination. Instead, what this blog is about is earning the right to be freed from such societal constraints. It’s about learning to self-regulate, so that, if and when mental storms pass through, they no longer require such harsh societal intervention.

I do not blame society for wanting to regulate unruly behavior. When people are seriously out of control, threatening themselves or others, or “acting weird” enough in public that observers feel the need to call 911, it can wreak havoc on civil society. Placing constraints on this type of behavior is the obvious thing to do in response.

So, the question for me is not: “How much blame should we place on our captors?” It is instead: “How can we control ourselves, such that captivity is no longer necessary?”

Captivity

By my count (with an assist from my mother) I’ve had 12 psychiatric hospitalizations in my life. I was first diagnosed with bipolar disorder in 1998. I had been in graduate school studying public policy at Georgetown University, and worked at the National Rehabilitation Hospital, which serves patients with traumatic brain injury, spinal cord injury, and other serious conditions. One of my assignments was to examine the string of acquisitions being made by the for-profit hospital chain HealthSouth. In 1998, I was the lead author in an article published in Health Affairs, the leading health policy journal, calling into question whether the rapid consolidation of the rehabilitation hospital industry was in the public interest. Later in that same year, I had a difficult break-up with a woman I considered to be my soulmate.

Shortly after the article was published, and the break-up occurred, I was hospitalized and diagnosed as bipolar. I was told I would be on psychiatric medications for the rest of my life. From 2001 to 2008, I was hospitalized for mania at a rate of almost once per year (7 times in 8 years)—including one month-long hospitalization. That lengthiest hospitalization occurred just prior to the Indian Ocean earthquake and tsunami of 2004, which killed 230,000 people. Whether my craziness was like the craziness that horses exhibit right before earthquakes hit, I cannot say. I do know that I was also hospitalized in 2005, just as Hurricane Katrina was getting ready to make landfall. My other episodes came during periods of personal crisis, when—as one psychologist later described it to me—mania served as a rocket ship, taking me far away from feelings of personal despair.

After being hospitalized in 1998, I was placed on lithium, and I took that drug for many years, until my kidneys began to show signs of damage. I was then placed on Depakote, which is the “mood stabilizer” I continue to take up until today. I have also taken the antipsychotic Zyprexa, first on an as-needed basis and now as an everyday commitment.

Efforts at Self-Regulation

Being placed in psychiatric hospitals at a rate of almost once per year was greatly disturbing, and it provided me with motivation to get my situation under control. So, I developed from scratch a mood tracking system that was designed to help me monitor my condition and promote self-regulation. This solution has proved to be enormously successful.

The system includes a series of 10 mood-related questions that I placed in an Excel spreadsheet (e.g., How is your mood right now? Have you been smiling/laughing? Are you feeling optimistic about the future?). There are two answer keys (one ranging from Great to Terrible, the second ranging from Yes!! to No!!). I administer these questions three times a day (at noon, 5pm, and 9pm), every day. On certain days, for example when my mood is particularly high, I complete the mood assessment more frequently.

The spreadsheet calculates a mood score, ranging from 0 to 100. Anything above 50 is an up mood, anything below 50 is a down mood. It is very difficult to score above 90 or below 10, because the exclamation points in one of the answer keys capture a level of emphasis that is rarely felt. Generating a mood score is very revealing; it enables me to identify specific things that are driving my mood higher or lower. It is not unlike the tools diabetics use to track their blood sugar. In using those devices, diabetics learn how specific foods and exercise affect their body. The tracking system I use does the same thing for my moods.

My highest mood score came on the day in 2009 when my only child was born. I was an 85 that day. However, I wasn’t at all manic at that point. So, I realized that I needed to create a second metric: not just a mood score, but a score that would specifically capture my mania level. (Other mood tracking systems consider elevated mood and mania risk to be one and the same thing. I believe this is inappropriate and actually detrimental.)

I decided to capture my mania scores on a scale of 0 to 10, using a traffic-light indicator of red, yellow, and green. What I wanted was a tool that would give me a red flashing light: an indication that I was becoming manic, and a warning that I needed to take steps to correct it.

I created a system where 0 to 5 represents a green light (not manic), 5 to 7 is a yellow light (caution), and 7 to 10 indicates a flashing red light. Over time, with self-reflection and extensive trial and error, I created a system that matches my mania level to a T. If the system says I am a 6.8 yellow, that represents a completely accurate depiction of my mania level at that point in time. I’ve tested the system in all kinds of situations, and it captures my mania level accurately on a consistent basis.

In designing the system, I added five mania-related questions, in addition to the 10 mood-related questions. First, how many hours of sleep did I get the night before? In the beginning, I gave considerable weight to this variable, given the amount of attention it always receives from doctors and in the literature. But, through trial and error, I found that I needed to lessen the weight of this variable considerably. On nights when I had not gotten a lot of sleep, my mania levels did not rise as much as might be expected. I had to continuously reduce the weight of this variable until it more accurately reflected the impact that sleep (or lack thereof) had on my mania. (Notice that just the process of creating the mania score enabled me to gain more awareness about the degree of mania I was experiencing at any given time.)

Second, I asked whether someone in my life (usually a family member or friend) had told me that I was acting manic. This variable, it turned out, needed to be weighted very heavily. It was weighted so heavily, in fact, that I found I needed to create a distinction that reflected the seriousness of the person’s expressed concern. Did they vaguely imply a level of concern? (In Excel I weighted this at 0.5). Or did they tell me directly, in no uncertain terms, “You seem manic” (this I weighted at 1.0).

Third, I captured “Big Ideas.” This reflected the number of ideas running through my mind that I considered thrilling. If I had three such ideas in the morning, I entered 3 into Excel. If I had one such idea, I entered 1. In a Word document, I wrote down what the idea was about. This Word document was actually an integral part of the system. Every day, during the time segments of morning, afternoon, and evening, I would capture the issues and events that were adding to, or subtracting from, my mood and mania scores. Writing these down allowed me to generate lessons learned that I could later use as tools for curbing my mania.

Fourth, I had a category called “Caught Manic Thoughts.” This reflected the number of thoughts I had that I knew to be manic or psychotic in nature (e.g., somebody is reading through my e-mails). I knew that this is the kind of thought I would be thinking if I was starting to become more manic, so I marked it down in Excel. The fifth and final category was “Believed Manic Thoughts.” These were manicky thoughts that I didn’t see as being the product of mania. Instead, I believed that somebody really was reading through my e-mails.

Obviously, it takes a certain degree of self-awareness to capture a Believed Manic Thought in Excel. Truth be told, this self-awareness is the therapy provided by my system. What is vital about the system is not only that it captures (numerically) the person’s mood and mania level during specific points of their life, but it cultivates self-awareness. It turns out, in fact, that self-awareness is an antidote to mania. It can make someone realize: I’m starting to become manic, and this is why, and this is what I can do about it.

I found that having the early warning system provided by the mood tracker enabled me to take steps to counteract the mania. For example, I learned that eating helps. Exercise helps. Talking to people helps. My therapist and I worked out an arrangement that, if I scored at or above an 8.0, and the system was flashing red, I would take the antipsychotic Zyprexa on a PRN basis. These steps helped me to remain out of the hospital.

I developed and used this system three times a day for many years, and in doing so, I substantially reduced both my emergency room utilization and my rate of psychiatric hospitalization. This has not only improved my life (because going to the hospital is not a pleasant experience), it has also resulted in direct savings of tens of thousands of dollars.

It was not just the numerical outputs that proved helpful, it was also the process of inputting the data. The system forced me to monitor my own thought processes. I became acutely aware of my manic thinking and then was able to take steps to counteract it. It was this process of self-regulation that allowed me to reduce the number of manic episodes I experienced—which reduced the number of emergency room visits and psychiatric hospitalizations I had to endure (at great expense to myself, my family, my friends, and the health care system overall).

Attempts to Break Free

I refer to this mood and mania tracking system as “The Grid.” This is in part because it captures energy levels (like a power grid). Scores above 50 are in the high energy range, scores below 50 are in the low energy range. It does not capture the broader rainbow of emotions (e.g., worried, anxious, disheartened, hopeful, peaceful).

I developed and actively used The Grid for a period of about five years (2008 to 2012). By the end of that period, I had internalized many of its lessons. I could easily estimate my scores even without using the system, and I was able to implement the lessons I had learned, demonstrating improved self-regulation. Though I continued to have hospitalizations, they occurred much less frequently. Maximizing the time in between hospitalizations is in fact a key indicator of success (and it obviously shows that the hospitalizations are declining).

Whereas from 2001 to 2008 I was hospitalized at a rate of almost once per year, from 2008 to 2013, I was hospitalized only once every three years. Based on my earlier trajectory, this meant that about four hospitalizations were avoided. I was hospitalized in 2013 in the context of my separation and divorce. I was also hospitalized in 2016. At that time, I was working in a challenging and, in some respects, toxic workplace. I suffered from stress and growing paranoia.

In fact, around this time I started to believe that my thoughts were being broadcast to those around me. Anything resembling an opinion running through my mind became an item of considerable controversy (though I had to keep up appearances that the thought had not become public knowledge). For months on end, I was in a “functional psychosis.” I believed that wars were taking place all around me and torture was rampant, yet I continued to function—more or less—as if nothing was wrong. I consider my ability to maintain functionality during this time to be an incredible feat of self-control—one that was made possible only because of the years I had spent doing The Grid, and being in therapy.

Around the 2016 hospitalization my diagnosis was elevated to schizoaffective disorder and I was placed on everyday Zyprexa. However, my paranoid thoughts began to subside shortly after I left my job in early 2018. Today I would say that my condition falls once again under the rubric of bipolar disorder with psychotic features (though today these features are rarely exhibited).

After 2016, my need for hospitalization declined. Whereas I had once been hospitalized nearly every year, by 2023 I was able to say that I had been hospitalized only once in the past 10 years, and in fact I had not been hospitalized at all in the past seven years. However, I was hospitalized again in December 2023 (for reasons I will explain below). So, in total, over the past 10-11 years, I can say that I’ve been hospitalized at a rate of approximately once every five years. This is still a considerable improvement over my initial trajectory.

By 2022, I had recovered to such a degree that I decided I wanted to come off the Depakote. I knew I could not give up Zyprexa without experiencing a sleep disruption, so I decided that Depakote was the better place to start.

My doctors supported this course of action mainly because I complained I was sleeping too much on my existing drug regimen. I had been averaging 9-12 hours of sleep a night. Though I had stopped using The Grid after 2012, I decided that if I was going to come off the Depakote, I would need to start using the system again. I have been using it basically every day since August of 2022 (almost two years).

One of the things I noticed in coming off the mood stabilizer was that my emotions moved closer to the surface. For example, I would begin to tear up at appropriate places while watching movies. I felt that this was an improvement in my emotional state, what I referred to as an “emotional thawing.” I also began to sleep less than I had when I was taking both drugs.

From my perspective, coming off the Depakote was successful. I went for almost a year and a half without a hospitalization. However, my therapist, who was looking out for me, expressed some reservations about it. During the 2022-2024 time period there were a number of stress factors in my life, in particular relating to finances. Since leaving my job in 2018 I have held only part-time jobs (though one of these was for a very prestigious organization). During this period, I have been drawing on savings, and this has caused financial stress. Through daily observation, what I now observe is that my mood scores rise and fall depending on how I’m feeling about finding a stable (and meaningful) work situation.

To address the ongoing stress that would arise in me, my doctors and I decided that I would go back on the mood stabilizer Depakote. In late November 2023, I took my first dose, moving from 0mg to 2000mg of Depakote in a single day. Within a week, while continuing to take the Depakote at 2000mg per day, I had a psychotic episode and was taken by ambulance to the hospital. I was admitted to a psychiatric hospital in December 2023 for a one-week stay. It seems clear to me that ramping up on Depakote so rapidly—in combination with the existing stressors in my life—is what caused this episode. Perhaps, I conjecture, the episode represented a fumbled hand-off between self-regulation and societal regulation. Following the hospitalization, I continued to take Depakote, but at a dose of 750mg per day. That is my current medication regimen (along with 10mg of Zyprexa per day).

Mood Tracking

There is no shortage of mood tracking systems available on the market. However, from my perspective as a patient, the tracking systems that are available through app stores and other sources are deeply flawed. Many are too general to be helpful. For example, some systems put the emoji of a smiley face on a calendar day, or they will ask the user to make a linkage between an emoji (e.g., “meh”) and some activity (e.g., “work” or “family”). These systems do not capture usable information and are ultimately unhelpful. Other systems generate numerical scores (e.g., 0% to 100%), but they provide no indication as to how to distinguish between 55% and 65%, for example.

The system I developed produces mood and mania scores based on simple questions (e.g., “Are you feeling enthusiastic/excited?”). The open text that is captured in Word allows the user to generate self-specific information, which they can use to develop strategies for achieving self-regulation. Once these skills are mastered, the need for societal constraints is dramatically reduced. Ultimately, this provides an avenue to reduce psychiatric hospitalization and other forms of patient restraint.