Patient-reported outcome after treatment for definite Lyme neuroborreliosis.

Abstract Objective To chart patient‐reported outcome measures (PROMs) in Norwegian patients treated for definite neuroborreliosis (NB). Material and Methods Adult patients treated for definite NB 1–10 years earlier supplied demographics, symptoms and treatment during NB, and answered validated questionnaires; Fatigue Severity Scale (FSS), Hospital Anxiety and Depression Scale (HADS), health‐related quality of life questionnaire (RAND‐36), and Patient Health Questionnaire (PHQ‐15). Results A higher proportion of NB‐treated persons reported severe fatigue, defined as FSS score ≥ 5, than in Norwegian normative data, but when removing persons with confounding fatigue associated comorbidities (n = 69) from the analyses, there was no difference between groups. Physical health‐related quality of life (RAND‐36 PCS), mean FSS score, proportions of persons reporting moderate or severe somatic symptom burden (PHQ‐15 score ≥ 10), anxiety (HADS‐A ≥ 8), or depression (HADS‐D ≥ 8) did not differ between NB‐treated persons and reference scores. Mental health‐related quality of life (RAND‐36 MCS) was poorer than in normative data (47.1 vs. 53.3), but associated with anxiety, depression and current moderate or severe somatic symptom burden, and not with NB characteristics. Conclusions Results on validated PROM questionnaires measuring fatigue, anxiety, depression, self‐reported somatic symptom burden, and physical health‐related quality did not differ between persons treated for definite NB 1–10 years earlier and reference scores. NB‐treated persons tended to report a slightly poorer mental health‐related quality of life than found in normative data, but when adjusting for confounders the causative connection is questionable. Overall, the long‐term prognosis of definite NB seems to be good.

the EFNS diagnostic guidelines (neurological symptoms suggestive of NB without other obvious reasons, pleocytosis in CSF and intrathecal Borrelia burgdorferi antibody production; Mygland et al., 2010) between 2006 and 2014. Twelve out of 18 invited hospitals agreed to participate. A contact person at each department reviewed the medical records for eligible patients (n = 598) and invited them per letter to fill in the PROM questionnaires, preferred web-based, but optional paper-based during 2015 to 2016.

| Variables and questionnaires
Year of NB diagnosis and CSF cell count were obtained from medical records.

| Patient-reported information about education level
Questions about education were obtained by reuse of formulations from the Nord-Trøndelag Health Study 3 (HUNT3), which is a Norwegian community-based survey (Krokstad et al., 2013).

| Symptoms
The patients were asked if they had experienced any of the following symptoms during their NB; Headache, fatigue, malaise, general pain, radiating pain from neck or back, facial palsy, paresis, numbness, memory or concentration problems, dizziness or unsteadiness, diplopia, or other. A sum score of symptom burden during NB was provided by giving each "yes" one point (max score 12).

| Treatment
The patients were asked if they were treated with orally or intravenously administered antibiotics or both, and for how many weeks Conclusions: Results on validated PROM questionnaires measuring fatigue, anxiety, depression, self-reported somatic symptom burden, and physical health-related quality did not differ between persons treated for definite NB 1-10 years earlier and reference scores. NB-treated persons tended to report a slightly poorer mental health-related quality of life than found in normative data, but when adjusting for confounders the causative connection is questionable. Overall, the long-term prognosis of definite NB seems to be good.

K E Y W O R D S
fatigue, neuroborreliosis, patient-reported outcome measures | 3 of 7 EIKELAND Et AL.
(by ticking off 1-6). For both questions, "I do not remember" was an optional answer.

| Patient-reported information about comorbidity
The patients were asked if they had ongoing or previous any of the following comorbidities: diabetes mellitus, metabolic disease, rheumatoid arthritis, fibromyalgia, systemic disease (sarcoidosis, systemic lupus erythematosus, or Sjögren's syndrome), other musculoskeletal disease, multiple sclerosis, Parkinson's disease, dementia, epilepsy, stroke, polyneuropathy, other neurological diseases, cancer, heart disease, asthma or chronic obstructive lung disease, allergy, chronic fatigue syndrome, depression or anxiety, other psychiatric disease, celiac disease, osteoporosis, kidney disease, other diseases not specified further. If they answered yes to one or more, the patients were classified as suffering from comorbidity. A sum score was provided by giving each "yes" one point (max score 24).

| Somatic symptom burden during the last 4 weeks
The Patient Health Questionnaire -15 (PHQ-15) is a validated questionnaire that charts prevalence and intensity of 15 somatic symptoms (stomach pain, back pain, pain in arms/legs/joints, menstrual cramps, or other problems with periods (women only), headache, chest pain, fainting spells, feeling of heartpounds or race, shortness of breath, pain or problems during sexual intercourse, constipation/loose bowels/diarrhea, nausea/gas/indigestion, feeling tired/having low energy, trouble sleeping) during the last 4 weeks (Kroenke, Spitzer, & Williams, 2002). The answers are graded "not bothered at all" (0 points), "bothered a little" (1 point), and "bothered a lot" (2 points). Sum score ranges from 0 to 28 for men and from 0 to 30 for women.
Scores 5-9 points, 10-14 points, and 15-30 points indicate mild, moderate, and severe somatic symptom burden, respectively (Zijlema et al., 2013). We translated the PHQ-15 questionnaire into Norwegian based on the Swedish version (Nordin, Palmquist, & Nordin, 2013). A missing value for a PHQ-15 item was replaced with the average value of the other items if the number of missing values did not exceed three items (n = 3; Kroenke et al., 2002).

| Anxiety and depression
Hospital Anxiety and Depression Scale (HADS) is a validated questionnaire measuring symptoms of anxiety and depression (Bjelland, Dahl, Haug, & Neckelmann, 2002). It includes seven anxiety items and seven depression items scored on a four-point Likert-like scale.
Scores ≥ 8 on the subscales indicate a need of further assessment of possible anxiety and/or depression (Leiknes, Dalsbo, & Siqveland, 2016). A missing value in either the anxiety or depression items was replaced with the average value of the other items if the number of missing values did not exceed one item in each subscore (n = 2).

| Health-related quality of life
RAND-36 is a validated questionnaire that measures health-related quality of life (Ware, 2000). It consists of 36 questions comprising eight multi-item scales: mental health, vitality, bodily pain, general health, social function, physical function, and physical and emotional role. These domains are then combined into physical and mental component summary scales (PCS and MCS).

| Statistical analysis
Results are given as proportions, mean with standard deviations (SD), or median with range as appropriate. To calculate 95% confidence intervals of differences, we used online calculators http://vassa rstats. net/prop2_ind.html (proportions) and https://www.socsc istat istics. com/confi dence inter val/Defau lt4.aspx (means). Selected variables, considered to be of importance, were entered into a multivariate linear regression analysis with mental health-related quality of life (MCS) score as the dependent variable. p-Values < .05 were considered statistically significant. All statistical analyzes were performed using SPSS version 25 (SPSS Inc.).

| Ethics
The study was approved by the Norwegian Regional Committees for Medical and Health Research Ethics, and all patients gave a written informed consent (REK 2013/1325).

| RE SULTS
Out of 598 invited NB patients, 307 gave a written informed consent to participate in the study, and 258 returned a satisfactorily completed questionnaire (228 web-based, 30 paper), yielding a response rate of 43%. Of ethical reasons (lack of informed consent), we could not obtain data from the nonresponders to detect meaningful differences between those who responded and those who did not.
Patient characteristics at the time of filling the questionnaire are shown in Table 1 was poorer among NB-treated patients than in normative data. To further analyze this finding, we did a linear regression and found that poorer mental health-related quality of life was associated with HADS-A ≥ 8 (p < .001), HADS-D ≥ 8 (p < .001), and PHQ-15 ≥ 10 (p = .024), but not with sex, age, years since NB diagnosis, education level, treatment administration route, treatment duration, CSF cell count at diagnosis, burden of symptoms at diagnosis, memory and/or concentration problems at diagnosis, comorbidity, or FSS ≥ 5.

| D ISCUSS I ON
The main results in the present study were that in our group of 258 Norwegian patients treated for definite NB one to ten years earlier the prevalence of severe fatigue, substantial symptomatic symptom burden, anxiety, and depression did not differ from relevant reference groups, and reported physical health-related quality of life was similar to normative data. Mental health-related quality of life tended to be slightly poorer among NB-treated patients, but was associated with anxiety, depression and presence of moderate or severe current subjective somatic symptom burden and not with any NB-related factors. Based on our findings, we therefore state that long-term prognosis of definite NB seems to be good.
Residual symptoms after treated NB is a much debated phenomenon (Halperin, 2017). Some have found residual symptoms in up to 50% (Eikeland et al., 2012;Eikeland et al., 2011;Knudtzen et al., 2017;Ljøstad & Mygland, 2010), while others have found that NB has no substantial effect on long-term survival, health, or educational and social functioning (Obel et al., 2018). Some of the disagreement may be explained by different study designs and poor case definitions (Dersch et al., 2016). TA B L E 1 Demographic features of the study population at the time of filling out the questionnaire